The genetic relationship between Alune, Lisabata, Luhu, and Wemale (Western Seram, Indonesia): a historical-comparative linguistics approach

Abstract

Language and cultural identity differences can easily lead to a small amount of warfare and confrontation due to a lack of sense of belonging. This study examined the genetic relationship of four languages spoken in Western Seram, consisting of the Alune, Lisabata, Luhu, and Wemale, and its urgency to the relation among four existing tribes by stepping on qualitative and quantitative evidence. To determine language kinship, 200 basic lexical items from Swadesh’s lists were tested on 12 informants. Along with that, 20 additional informants were added for face-to-face interviews to assess historical, philosophical, social, and cultural backgrounds. Despite the inconsistencies in phonological change regularities, the results have revealed that Lisabata and Luhu had the highest percentage of language kinship at 64% with 128 cognates, followed by Lisabata and Alune at 58% with 116 cognates, Luhu and Alune at 57.5% with 114 cognates, Alune and Wemale at 49% with 98 cognates, Wemale and Luhu at 47.5% with 95 cognates, and Wemale and Lisabata at 45% with 90 cognates. Evidence of interconnectedness between the four languages on the family level within the historical frame of Nunusaku mythology can be used further as a milestone to strengthen the relationship between the four ethnic groups.

Keywords:

• Ethnic identity
• genetic relationship
• historical-comparative linguistic
• language kinship
• Western Seram

REVIEWING EDITOR:

• Jeroen van de Weijer, College of International Studies, Shenzen University, China

SUBJECTS:

• Culture
• Race & Ethnicity
• Historical & Comparative Linguistics
• Interdisciplinary Language Studies
• Linguistic Theory
  

1. Introduction

Cultural identity, which is identified as a fluid process and a broad set of identity constructs shaped by historical narratives, social norms, and collective memories (Amiot et al., 2018; Pande & Jha, 2016; Worrell, 2020), often transforms into a battleground where opposing worldviews collide. This primary justification lies in the intricate interplay between the human need for a sense of belonging and the inherent diversity that characterizes a society with many different cultural backgrounds. As individuals and groups navigate their identities in an ever-evolving world (Lustig, 2013; Pande & Jha, 2016), the quest to preserve one’s cultural heritage can sometimes lead to a defensive stance that inadvertently deepens the fault lines between communities. The deep-rooted concept of ‘us (as the representation of “self”) versus them (as the representation of “others”)’ (He, 1995; Jenkins, 2008; Sekulić, 2008; Peoples & Bailey, 2011) creates boundaries and drives wedges between individuals who share common values and aspirations but are divided by the nuances of their own cultural expressions.

When discussing culture, it is important to recognize that cultural identity is inextricably linked to sociocultural factors such as language, societal norms, history, and belief systems. Some factors can be more salient and intense than others. Among those four factors, the association between language and identity is profound and intricate. Since culture is composed of explicit and implicit patterns of and for behavior, which are acquired and transmitted by symbols and communicated adaptively through language (Kroeber & Kluckhohn, 1952; Weitzman, 2013), language appears as a pivotal marker as well as a powerful engine of cultural affiliation and ethnic identity (Dorian, 1980; Fishman, 1996; Gelman & Roberts, 2017; Lytra, 2016; Reyes, 2010; Villegas-Torres & Mora-Pablo, 2018). Language, as a supporting system for unlimited heredity of cultural information (Gelman & Roberts, 2017), is essential for forming, expressing, and negotiating individual and group identities. It also serves as a medium through which individuals convey their cultural affiliations, personal histories, and social sense of being part of a particular ethnic group. Thus, as proposed by Malinowski (1961), language is strongly influenced by culture, so everything contained in the culture will be reflected through language.

At the most prominent level, ethnic groups are characterized by a range of interconnected factors, such as genetic ancestry, cultural heritage, history, homeland, language, dialect, religion, mythology, folklore, ritual, cuisine, dressing style, art, or physical appearance, that contribute to their distinct identity and shared senses of belonging (Ananta et al., 2015; Baumann, 2004; Cohen, 2004; Epstein & Heizler, 2015; Gong, 2007; Green et al., 2015; Phinney & Ong, 2007; Radcliffe & Pequeño, 2010). These factors encompass both tangible and intangible aspects that shape how individuals within the group perceive themselves and how others perceive them. Language acted as a symbol of the group’s uniqueness, as well as of the group’s cultural heritage. Each ethnic group often has distinct languages or dialects that act as instruments of self-manifestation and inter-ethnic communication. Therefore, language constituted the storehouse of ethnicity, with each ethnic group expressing and identifying itself by the language it spoke.

Although Keyes (1981) notes that the study of ethnicity is theoretically impasse, there has been an increasing interest in ethnic group research, which can be shown through the growing number of substantial studies by viewing ethnicity as a multifaceted social construct that influences both personal identity and group social relations (M. E. Ford & Kelly, 2005; C. L. Ford & Harawa, 2010). Observations were made in numerous situations and approaches. Guided by social constructionists’ and anthropologists’ perspectives, Antweiler (2015) defines ethnicity as both a cognitive-cum-emotional and behavioral phenomenon that depends on place, time, situation, and context. He argued that ethnicity is dynamic and dependent upon context rather than static since some people’s ethnic identities and affiliations may change due to intermarriage, cultural exchange, political change, migration, and assimilation. On the same basis of argument, according Nagy et al. (2014) and Noels (2014) ethnicity is a loosely connected cluster of thoughts, feelings, and behaviors shared by members toward their ancestral ethnic group. In turn, different aspects of personalities can affect peoples’ stereotypes toward their ethnic group.

However, due to numerous psychological, sociopolitical, and cognitive convergences, language and identity have potentially triggered social division and conflict that causes ethnic or group separation (Bonomi et al., 2021; Charness & Chen, 2020; Gartzke & Gleditsch, 2006). Driven by the ‘homeland security’ agenda, the imposition of dominant cultural narratives has the potential to marginalize minority identities while also fostering feelings of exclusion and alienation. Several key factors contribute to this unforeseen event, such as identity threat and insecurity, identity and boundary maintenance, confirmation bias and cognitive dissonance, different interpretations of history and memory, fear of cultural homogenization, the political instrumentalization of identity, and the enduring impact of framing or stereotyping.

Conflict of interest within ethnic groups, cultures, and languages intensely occurs in societies. These conflicts probably arise from various circumstances and manifest in multiple ways that impact social cohesion, governance, and overall harmony. In some cases, most societies identify themselves as ‘different’ by separating between major (dominant) and minor (periphery) groups, between insiders and outsiders or mainlanders. In this circumstance, ethnic conflict occurs when ethnic groups compete for the same goal, including power, resource access, or territory. Individuals or groups may shift from one ethnic group to another through language shift, acculturation, adoption, and religious conversion. As a direct impact, ethnic groups may be divided into subgroups or tribes due to endogamy or physical isolation from their ‘immediate ancestors’.

Known as a multicultural country with a diverse range of races, ethnicities, and cultures, Indonesia is ripe for multi-dimensional conflicts due to its cultural diversity. Severe conflicts caused by primordialism (i.e. excessive attachment to subnational cultures, such as ethnicity, religion, race, regionalism, and family), ethnocentrism (i.e. assessment and judgment towards other cultures based on their own social values and cultural standards), and fanaticism (i.e. excessive obsession with certain ethnicities, cultures, or religions) have potentially occurred between various groups based on ethnicity, religion, or social position. Wealthy in traditional values, frequent tensions are likely to arise as a result of political rivalry over land and natural resources, sticking to the general convention that the genealogical position occupied in a clan determines the right to land use held in customary ownership. Since the violence in each place began with small-scale clashes between community groups but escalated into much larger confrontations, these tensions and feelings can further breed resentment, mistrust, and a sense of injustice when left unaddressed.

Given that ethnic divisions and conflicts are considered vulnerable, this study seeks to highlight the contribution of language in inter-ethnic relations. This study was centered around the Moluccas ethnic and language-speaking groups in Western Seram (Patasiwa Alfoeran), consisting of the Alune, Lisabata, Luhu, and Wemale tribes, as shown in Map 1.

Map 1. The territorial extent of the Western Seram Regency.

With the village functioning as the most important unit, the political situation of Western Seram displayed a great deal of particularism (Knaap, 1993). Due to population density and regional boundaries, conflicts over territory and/or personal animosities between village chiefs, as leaders of the seniormost clan in the settlement, or other prominent persons might easily result in a small amount of warfare and confrontation (Knaap, 1993). By considering the potential problems, this study tends to give brief overviews on: (1) How is the linguistic kinship of the Alune, Lisabata, Luhu, and Wemale tribes examined using quantitative and qualitative approaches based on historical-comparative linguistic theory?; and (2) What is the importance of providing historical-comparative linguistic evidence regarding the kinship of the Alune, Lisabata, Luhu, and Wemale tribes?

Considering some languages within the same or adjacent region potentially have similarities as they originated from the same proto (Sanjoko & Erniati, 2020), some scholars have already conducted intensive studies on Western Seram ethnic groups. By giving considerable attention to the existence of the Lisabata language, Erniati (2018) formulated Lisabata syllable patterns and split them into V, VK, KV, KVK, VKV, KKVK, ½ KV patterns (where V stands for vocal ‘vokal’, while K stands for konsonan ‘consonants’). Erniati and Sanjoko (2020) investigated Wemale syllable patterns by positioning Wemale as an endangered language. They then categorized Wemale syllables into V, KV, VKV, KVV, and ½ KV. Further, Simon (2015) and Darman (2022) explored the Alunes’ and Wemales’ kinship. They both concluded that the Alune and Wemale languages were family-related, viewing from the number of word similarities used by both languages. Meanwhile, Pattiasina (2021) highlighted and emphasized that the Alune language had become endangered time by time as the younger generation kept reducing the use of the language in their daily conversation.

Even though there were numerous examples of studies of Western Seram ethnic groups, all the studies mentioned earlier showed tendencies to only investigate language regularities. Nevertheless, since the modern-day scientific study of linguistics considers many aspects of language, further questions arose regarding the implication of such regularities. The evolution of each ethnic group and domain can be traced through language topologies. Because of that, language (like genes) is then properly understood as an ‘organism’ and ‘document of history’ (N. Evans, 2009; Gray et al., 2011). It can be said that local languages will remain alive as long as they serve functions in a particular community, whether in the domestic or public sphere (Sedyawati, 2008). The study of genetic linguistics leads naturally to the study of anthropology and archaeology by concentrating on language change and reconstruction (Antilla, 1989; Gray et al., 2011). To this extent, the contribution of historical-comparative linguistics combined with the in-depth discussion beyond identity and ethnicity is necessary as the apparatus to preserve the existence of endangered languages and reunite the pluralistic society in the Western Seram Regency area.

2. Literature review

2.1. Ethnicity and tribal affiliation as sources of collective identity and political allegiance

Ethnicity and tribal affiliation or community attachment, as a form of collective identity firmly rooted in history, provide people with a sense of continuity and temporal depth (Muldoon et al., 2017; Smeekes & Verkuyten, 2015; Van Bergen et al., 2015). While collective identity continuity’s feelings positively impact group processes and intergroup relations, identity continuity’s perceptions can negatively affect intergroup attitudes when people have an exclusionary understanding of their group identity (Roccas & Elster, 2012; Smeekes & Verkuyten, 2015). Beyond this concern, inter-ethnic conflicts, such as the ramifications of ethnic differences and inter-ethnic competition over resources and rights, manifest multi-dimensional causes of primordial social exclusion.

Since identity is multifaceted, containing at least two aspects: the internal and external dialectic of identity (Michel & Ben-Slimane, 2021; Peng et al., 2020; Zhang, 2023), identity is not only a matter of self-awareness and belonging but is also partially shaped by recognition from others (Taylor, 1992; Stets & Burke, 2003; Zhang, 2023). In another way, if ethnicity confirms the existence of a group identity, then according to Lowe and Tsang (2017), there must be ‘confirmation of sanctioned behavior within the group and demonstration that group norms can be observed’. The terms indigenous people and tribal highlight various historical and social contexts structured by hierarchical power. Once there are one or more dominant viewpoints, it is religious affiliation, political allegiance, national identity, and ‘differences between natives and settlers’ that overlap and reinforce each other (McBride, 2023). Each party controls its own semi-independent interest groups utilized for policy purposes within a relationship based on ethnic identification (J. Evans & Tonge, 2013; Kapidžić, 2019). Because of that, the appeal of ‘ethnicity’ to ethnic conflict tends to be descriptive rather than causal, although the dividing line between them was rarely clear-cut.

2.2. A general understanding of the Alune, Lisabata, Luhu, and Wemale

As Collins (1982) proposed, all the indigenous languages of Western Seram appear to be descendants of a single ancestral language, labelled elsewhere as ‘Nunusaku’. The term is derived from the name of one of the kingdoms in the western Seram mountains, which is thought to be the primordial site and the genesis of life on Seram Island and even the Moluccas region. The present-day languages in Western Seram can be categorized into two branches of Nunusaku: one stemming from the Tiga Sungai Purba ‘Three Ancient Rivers’ (mostly known as Saniri Waélé Têlu ‘Council of Three Rivers’, Tiga Sungai Besar ‘Three Great Rivers’ or Tiga Batang Air ‘Three Trunks of Water’) and the other from the Proto-Piru Bay or ‘Teluk Piru Purba’. Based on these historical frameworks, Amalumute groups (e.g. Alune, Lisabata, and Luhu) and Wemale are detected to belong to the same cultural frame of the ‘Three Rivers’ (refers to Eti, Tala, and Sapalewa rivers), as shown in Map 2.

Map 2. Groups of languages spoken on the Western Seram (Collins, 2018, p. 98).

The Austronesian language family or Proto-Austronesian, covers all the languages and dialects spoken by Western Seram tribes. The Alunes (also known as Sapalewa or Sapolewa people) are one of the long-established ethnic groups of Seram Island. As part of Indonesian ethnic groups family, in their daily communication basis, the Alunes speak Malayo-Polynesian, with Kairatu, Central West Alune (e.g. Niniari, Lumoli, Piru, Riring), South Alune (e.g. Rambatu, Manusa, Rumberu), North Coastal Alune (e.g. Nikulkan, Murnaten, Wakolo), and Central East Alune (e.g. Buriah, Weth, and Laturake), as their dialects (Alune—Peta Bahasa, n.d.; Asher et al., 2018; Frawley, 2003; Grimes, 1992; Sanjoko & Erniati, 2020). Likewise, the Malayo-Polynesian language is also spoken by northern Wemales (with Horale, Kasieh, and Uwenpantai dialects) and southern Wemales (with Waraloin, Honitetu, and Kawe dialects) (Asher et al., 2018; Frawley, 2003; Grimes, 1992; Sanjoko & Erniati, 2020; Wemale—Peta Bahasa, n.d.); Lisabata community (spoken by Lisabata native speakers in East Lisabata, Nualiali, Sukaraja, and Kawa (Asher et al., 2018; Frawley, 2003; Grimes, 1992; Erniati, 2018); as well as Luhu community (spoken by Luhu native speakers in Luhutuban, North Boano, Waesala, Kaibobu and Kamarian (Asher et al., 2018; Frawley, 2003; Grimes, 1992; Luhu—Peta Bahasa, n.d.)). Consider Figure 1 below.

Figure 1. Subgrouping of ‘Three Rivers’ in Western Seram (Collins, 1982, p. 37).

According to Western Seram’s local tradition, Wemale originated from ‘Wai’, signifying ‘river’ and ‘Mala’ as the name of the southeast area of Western Seram (Knaap, 1993), while Alune derived from the word ‘Alune’, which means ‘mountain people’. Although each ethnic group of Western Seram had its own language and several distinctive cultural features, the intricate connection between Alune, Lisabata, Luhu, and Wemale within the ‘Three Rivers’ cultural tapestry (as shown in Figure 1) strongly reflects shared heritages, language typologies, and typical historical trajectories. Otherwise, Luhu was unique because, different to other languages, the language spoken by Luhu is a descendant of Proto-Piru Bay, as evidenced by its ties to Alune and Lisabata. Language becomes a ‘special identifier’ owned by each tribe. In this essence, Three Rivers’ existence has served as physical landmarks and vital cultural exchange and interaction conduits among tribes or ethnic communities.

2.3. Interlanguage and inter-ethnic correlation from the historical-comparative linguistics perspective

The comparative method is the most common tool linguists use to determine the relation between two or more languages (List et al., 2018; Power, 2022; Reagan, 2021; Weise, 2018). This standardized method provides insights into the historical development of languages by comparing lexical items, phonology, and grammar across languages to identify and comprehend the changes that occur over time (Antilla, 1989; Crowley & Bowern, 2010; Hock & Joseph, 2019; Lass, 1997; Reagan, 2021). Linguists have developed a meticulous set of procedures known as the ‘comparative method’ to infer ancestral states and construct language family trees by concerning with the reconstruction of an earlier state of a language based on comparisons of related words and expressions in different languages or dialects derived from it (Campbell & Poser, 2008; Durie & Ross, 1996). This internal reconstruction includes proto-language reconstruction, classification of languages into families, and subgrouping of more closely related languages within those families (Power, 2022), as generated in Figure 2.

Figure 2. The construction of Genetic tree model (Schleicher, 1983).

At the most critical point, historical-comparative linguistics approach provides a starting point and fair threshold for deriving the histories of all daughter (descendant) languages into the mother language as the root (proto) of the language descended through the genetic descent process. The overall syntheses and plausible intuitions were done by tracing their evolutionary history based on their ancestral lexical item and relative cultural chronology (Lass, 1997; Power, 2022). From a historical-comparative linguistics perspective, linguistic facts in the form of regularity found in related languages can be used to establish language kinship and origin through language pattern regularities. These genetic relatedness among languages can be used later as evidence of shared authenticity inherited from the same ancestor. Relating to what has been mentioned, Mortarino (2009), as well as Crowley and Bowern (2010), argue that when nearby languages are compared, linguists rely on complex information (such as sound correspondences, grammar rules, etc.). Lexicon comparison between two or more languages, resulting in systematic similarities and a series of sound correspondences originating from the same original sound.

3. Methods

3.1. Data sources and data collection techniques

This study adopted 200 basic lexical items from Swadesh’s Swadesh (1952) lists as sample entries. These classic compilations include terms for pronouns, numerals, certain body parts, geographical features or natural objects, and basic actions and states (Antilla, 1989; Gudschinsky, 1956; Gray et al., 2011). However, because words like bark (of a tree), freeze, ice, leg, and snow were exotic and belonged to culturally specific items, following Keraf’s Keraf (1996) theory, they were replaced with the moon, fingernail, breast, wing, and knee. Since no similar items exist in most Austronesian languages and cultures (in this case, Indonesia only recognized ‘kaki’ represented by the word foot and did not recognize snow due to the absence of winter season), these substitutions were prevalent to prevent untranslatability or referential vacuum.

By conducting parallel face-to-face interviews from December 2022 to January 2023, 12 native Alune, Lisabata, Luhu, and Wemale speakers, with equal ratios, were asked to mention the words included in the Swadesh’s Swadesh (1952) lists in their native languages (see Figure 3). Because this study is concerned with ethnicity and identity, 20 additional informants (consisting of 5 informants from each region) were added to collect historical, philosophical, social, and cultural backgrounds, which can be used as supplementary data. All of the informants were selected using snowball sampling (i.e. a recruitment technique in which informants in the study were asked to assist researchers in identifying other eligible subjects (Vanderstoep & Johnston, 2009) based on several criteria, such as (1) indigenous residents; (2) between the ages of 30 and 50; (3) recommended by tribal council; (4) having no language and hearing impairment for ensuring clarity; as well as (5) having limited mobility (never live outside the Alune, Lisabata, Luhu, and Wemale areas for extended periods), as well as having sufficient education levels for maintaining the authenticity and purity of the spoken languages. All collected data were then manually transcripted and organized into table lists to accommodate the comparison process’s effectiveness. Finally, lexical cognates were identified based on recurrent sound correspondences by (a) assembling comparable word lists from the relatively stable core lexical item by stepping on the historical-comparative linguistics approach; and (b) determining the probable cognates of Alune, Lisabata, Luhu, and Wemale languages. The whole related words (cognates) and unrelated words are presented in Online Appendix 1.

3.2. Data analysis techniques

This diachronic study (i.e. the study of language by establishing the facts and degree of relationship for two or more languages (Antilla, 1989)) implemented a sequential mixed-methods design by integrating quantitative calculation and qualitative analysis. The overall data were separated based on their relative and non-relative words. Kinship statuses between Alune, Lisabata, Luhu, and Wemale were obtained from the number of relative (homologous) words or counted lexicons. Under this hypothesis, pairs of words would be justified to be relative or judged to be cognates if they were identical, phonemically correspond, phonetically resemble, and have at least one different phoneme (Keraf, 1996). Lexicostatistical tests were subsequently employed in this study as a method of grouping languages by following these steps: (1) eliminating unrelated words; (2) determining cognacies, which were decided through strict observance of meanings (Antilla, 1989); (3) calculating the number of cognates (cognate set); (4) determining the percentage of cognates; and (5) comparing cognates to postulate or reconstruct the proto-language. Discovered by Swadesh (1952, 1955), Formula 1 was used in this study to calculate the percentage of cognates. (1) $$\mathit{\text{\%\hspace{0.33em}of\hspace{0.33em}}}\mathit{cognates}(C_1)=\sum \frac{\mathit{\text{cognates\hspace{0.33em}\hspace{0.33em}or\hspace{0.33em}\hspace{0.33em}kindship\hspace{0.33em}\hspace{0.33em}}}\mathit{languages}}{\mathit{\text{basic\hspace{0.33em}\hspace{0.33em}vocabularies\hspace{0.33em}\hspace{0.33em}or\hspace{0.33em}\hspace{0.33em}}}\mathit{glosses}}\times 100\text{\%}$$(1)

The following parameter assessed the level of language kinship used to indicate the degree of kinship in the proto-language. The parameters for determining language kinship are presented in Table 1.

Table 1. Language kinship parameter (Keraf, 1996, p. 135).

Levels	Split times (in millenia or centuries)	Cognate percentages (%)
Language	0–5	100–81
Family	5–25	81–36
Stock/Phylum	25–50	36–12
Microphylum	50–75	12–4
Mesophylum	75–100	4–1
Macrophylum	100 and above	1 less than 1

Genetic linguistics has its own decay dating, known as glottochronology (Antilla, 1989). It was the extension of lexicostatistics that rested upon the statistical comparison of the basic lexical item shared by two or more related languages based on the assumption that the rate of lexical item replacement is constant over sufficiently long periods of time (Gray et al., 2011; Lees, 1953). Formula 2 was used to calculate the length of time’s separation (split times), where t₁ belongs to the time depth in millennia, C₁ belongs to the percentage of cognates shared by languages, while r belongs to the ‘universal’ constant or retention rate (the expected proportion of cognates remaining after 1000 years of separation). (2) $$t_1=\frac{\mathrm{\text{log}}\hspace{0.17em}{C}_1}{2\mathrm{\text{log}}\hspace{0.17em}r}$$(2)

Nonetheless, as the calculation of separation time may contain errors, formula 3 was used for calculating the standard deviations (S₁), with C₁ representing the percentage of cognates and n for the number of comparable words (without exception). (3) $$S_1=\frac{\sqrt{C_1(1-C_1)}}{n}$$(3)

In order to obtain recent or new cognates percentage calculation (C₂) results, the results of the standard deviation (S₁) calculation are then summed with the percentage of the ancient cognates (C₁) by using formula 4. (4) $$C_2=C_1+S_1$$(4)

The new length of time separation (t₂) or time depth in millenia was later recalculated using formula 5. Meanwhile, the standard deviations (S₂) and the separation range (t_beginningto t_end) were counted using formulae 6 and 7. (5) $$t_2=\frac{\mathrm{\text{log}}\hspace{0.17em}{C}_2}{2\mathrm{\text{log}}\hspace{0.17em}r}$$(5) (6) $$S_2=t_1-t_2$$(6)

$t_{\mathit{\text{beginning}}}=t_1+S_2$ to $t_{\mathit{\text{end}}}=t_1-S_2$(7)

After completing all the stages, researchers validated the results by assessing the similarity between quantitative and qualitative evidence. An in-depth examination of complete sound correspondences became the final step before discussing how language kinship becomes a medium for inter-ethnic reconciliation. In this case, sound correspondence is the parallelism of sound in the same position in derived languages based on the collection of basic words. This parallel can be seen through the intensity of changes in sound and meaning that occur regularly in four languages. In this study, sound correspondences were identified by considering these general principles: (a) any reconstruction should involve sound changes that are plausible unless there is good evidence to the contrary, and (b) any reconstruction should involve as few changes as possible between the proto-language and the daughter languages by turning attention to the consonant correspondences (in which the daughter languages all have the same reflex) and sound correspondence sets that only have slight differences between the various daughter languages.

4. Findings and discussion

This section delves into the Alune, Lisabata, Luhu, and Wemale language kinship, which proceeded according to the rules of language reconstruction as a methodological procedure and extensive validation to determine the characteristics of an unattested ancestor language of one or more given languages. These findings addressed the first research question by involving quantitative calculations (identified in sections 4.1, 4.2, and 4.3) and qualitative findings (presented in section 4.4). Meanwhile, the explanation regarding the second research question was raised in section 4.5.

4.1. The determination of the Alune, Lisabata, Luhu, and Wemale language genealogy through language kinship identification

The Alune, Lisabata, and Luhu languages have produced numerous correspondence opportunities due to their minimal changes in phonological forms for the same meanings. The distribution of sounds in four existing languages occurred elsewhere, either in the initial, medial, or ultima/final positions of native lexical items. The data presented in Table 2 consisted of representative data that represented all the cognates found in this study.

Table 2. Sets of language kinship between the Alune, Lisabata, Luhu, and Wemale.

No	Gloss	Alune	Lisabata	Luhu	Wemale
1.	I	au	a’u	a’une	yau
2.	wind	-	anin	anine	yaline
3.	dog	asu	asu	asu	yasu
4.	you	ale	ane	alene	yale
5.	dig	kali	kasi	kasi	ali
6.	nose	inubata	inu	ninu	niru
7.	wife	bina	mahina	mahinae	dapina

Despite inconsistencies in phonological change regularities (looking at pronunciation and sound system structure differences), the similarities and differences between the Alune, Lisabata, Luhu, and Wemale seem obvious since possible sounds may be changed, merged, lost, and split. Through time, language creates a new system of phoneme oppositions. As the impact, old contrasts may disappear, new ones may emerge, or they may be rearranged.

On the border between the Alune, Lisabata, Luhu, and Wemale, Table 2 clearly shows variation compatibility marked by the existence of zero phonemes [Ø] at the initial of the Alune, Lisabata, and Luhu and the existence of phoneme [y] at the initial of the Wemale for glosses ‘I’, ‘wind’, ‘dog’, and ‘you’ (as shown in data 1 to 4), as well as addition and deletion in various parts. For instance, in datum 1, the extensive phoneme changes happened in the set of [au]-[a’u]-[a’une]-[yau] for gloss ‘I’ by adding primary stress to phoneme [a] in Lisabata and Luhu, adding phoneme [e] in the final position of Luhu, and inserting phoneme [y] in the initial position of Wemale. Within the same paradigm, in datum 2, extensive phoneme changes happened in the set of [Ø]-[anin]-[anine]-[yaline] for ‘wind’. The exact rules were also found in other glosses, such as in the gloss ‘dog’ (datum 3), with [asu] as the identical set of phonemes used by the Alune, Lisabata, and Luhu, and [yasu] used by Wemale; as well as in ‘dig’ (datum 5) with [kasi] as the identical set of phonemes used by Lisabata and Luhu; and [ali] used by Wemale (with the presence of zero phonemes in initial position and substitution of dental-ficative [s] into dental-lateral [l] in the middle position). In some exceptional cases, even though there were quite striking differences in data 6 and 7, as there were significant changes from [inubata]-[inu]-[ninu]-[niru] for ‘nose’ and [bina]-[mahina]-[mahinae]-[dapina] for ‘wife’, these patterns still demonstrated related words, as indicated by the appearance of several identical phonemes in both data. Overall, the findings supported Keraf’s Keraf (1996) argument that sets of words were considered relative or cognates if they were identical, phonemically corresponding, phonetically resembled, or had at least one different phoneme.

Several cognates were also derived from identical words with identical orthographic distributions and representations. It means all parts of the lexeme (phonological/orthographic form) were already in place. Consider Table 3.

Table 3. Sets of identical words shown by the Alune, Lisabata, Luhu, and Wemale.

No	Gloss	Alune	Lisabata	Luhu	Wemale
1	father	ama	ama	ama	ama
2	mountain	ulate	ulate	ulate	ulate
3	mother	ina	ina	ina	ina
4	louse	utu	utu	utu	utu
5	die	mata	mata	mata	mata

Zero phonemic changes in [ama] for ‘father’, [ulate] for ‘mountain’, [ina] for ‘mother’, [utu] for ‘louse’, and [mata] for ‘die’, became substantial evidence that four languages were intricately linked and rooted from the same ancient language. This discovery opened a fascinating window into linguistic evolution, shedding light on the deep historical connections that bound them together over millennia.

Nonetheless, many continuums, such as geographical location, phonological distances, and separation length from the ancient language, have contributed to the changes and shifts from the ancient to the language used today (Burenhult, 2008; Hock & Joseph, 2019). To this extent, despite the comparison between Alune, Lisabata, Luhu, and Wemale languages produced similar and identic words, it also showed many unrelated word pairs, as shown in Table 4.

Table 4. Sets of different words shown by the Alune, Lisabata, Luhu, and Wemale.

No	Gloss	Alune	Lisabata	Luhu	Wemale
1	worm	tilatinɛ	lɛlaɛ	wale	latupote
2	and	kai	te	duwe	tuna
3	near	tɛyanɛ	masu	waiway	mati
4	dirty	nalɔpɔ	cammarɛ	asasini	ma?inini
5	short	rɔkɔnɛ	noto	etey	tupe?i
6	think	nɔwakɛ	alae?e	hanaluwe	pikir
7	a little	ɛputi	sila?a	lahune	toto?o

The data set on Table 4. shows that there are no pronunciation similarities in [tilatinɛ]-[lɛlaɛ]-[wale]-[latupote] for ‘worm’, [kai]-[te]-[duwe]-[tuna] for conjunction ‘and’, [tɛyanɛ]-[masu]-[waiway]-[mati] for preposition ‘near’, [nalopo]-[cammarɛ]-[asasini]-[ma?inini] for ‘dirty’, [rɔkɔnɛ]-[noto]-[etey]-[tupe?i] for ‘short’, [nɔwakɛ]-[alae?e]-[hanaluwe]-[pikir] for ‘think’, as well as in [ɛputi]-[sila?a]-[lahune]-[toto?o] for ‘a little’. According to François (2015) and Markov et al. (2023), languages that come from the exact origin can also undergo changes that lead to the emergence of language variation due to various selective pressures. François (2015, p. 161) then added that over a sufficiently long period, changes in a language can accumulate to such an extent that it is no longer recognizable as the same language. This situation also happened in the Western Seram languages. Population migration, which occupies areas with varying geographical conditions and distances, became one of the factors that influence such variations.

Through direct field observations, it was evidenced that the majority of the Alune and Wemale ethnic groups tend to live in mountainous areas, while the majority of the Lisabata and Luhu ethnic groups tend to live in coastal areas. As stated by Knaap (1993, p. 254), Western Seram traditional mountain people have become representatives of a primitive sort of society or hunter-gatherers, swidden agriculturists, and head-hunters with restricted access to transportation and education, since the 1650s. In contrast, the coastal villages are predominantly inhabited by fishermen and traders who have migrated from the mountainous interior to the coastal areas (Knaap, 1993, p. 254). Thus, linguistic contact can lead to the development of similar and identical words, as observed in the Lisabata and Luhu languages, or dissimilar words, as evident in the Wemale language.

4.2. Beyond the numbers: the determination of the Alune, Lisabata, Luhu, and Wemale language genealogy through lexicostatistical calculations

In this study, the emergence of a column from four juxtaposed languages provides a ‘lexicostatistical’ measure of each language’s conservatism for the considered features. The ratio of these results resulted in the average retention rate for the given features, as shown in Table 5.

Table 5. The percentage of language kinship (cognates) among Alune, Lisabata, Luhu, and Wemale.

	Alune	Wemale	Lisabata	Luhu
Alune		49%	58%	57%
Wemale			45%	47,5%
Lisabata				64%
Luhu

Shading shows that there is no comparison between same languages 

The quantitative evidence toward 641 cognates derived from 200 Swadesh lexical items has revealed that Lisabata and Luhu had the highest percentage of language kinship at 64% with 128 cognates, followed by Lisabata and Alune at 58% with 116 cognates, Luhu and Alune at 57.5% with 114 cognates, Alune and Wemale at 49% with 98 cognates, Wemale and Luhu at 47.5% with 95 cognates, and Wemale and Lisabata at 45% with 90 cognates. Based on the results of the lexicostatistical classification and various extra-linguistic evidence, the Western Seram language family is then depicted in Figure 4.

Figure 3. Situation encountered during interviews with indigenous residents as respondents (researchers’ self-documentation, 2022).

Given the profound influence of religion on any society or culture (Balraj et al., 2020; Bin, 2023; Edara, 2017), the role of religion in tribal culture naturally encourages interaction between Luhu and Lisabata. The onset and development of Islamization became a vital entity in Luhu and Lisabata’s historical development and civilization. Occupied by Ternate and Tidore Sultanate around the 13^th-14^th century, Luhu was believed to be the epicenter of Islam’s expansion in the Hoamoal Kingdom, evidenced by the existence of the Batu Kapal site, Pulau Buano site, Jami Luhu mosque, ancient Qur’an, ancient graves, and Islamic-style artifacts (Handoko, 2016, pp. 102–110). Luhu religion then extended to the Lisabata (Sapalewa) community, substantiated by its distinction as the largest Muslim population in Taniwel. By this time, religion as the catalyst in establishing intra-community solidarities then creates social boundaries, leading to language contact among communities (Adamou & Matras, 2021; Spolsky, 2003).

The intersection of Lisabata and Alune is related to their origins and ethnic affinities. Based on the interview data, Lisabata is among the communities responsible for safeguarding the Three Trunks of Water (Kwélé Batai Têlu), comprising the Eti, Tala, and Sapalewa rivers. These water flows serve as the lifeblood of the Western Seram community, spanning the entire territory of Alune. It comprises four hierarchies: Inama, which stands for Nuniali; Pati, which stands for Lisabata; Angkota, which stands for Wakolo; and Sailbubui, which stands for Buria. The four groups of Alune continue to be upheld to this day. Hence, Lisabata maintains a strong connection with Alune.

Alune and Luhu share a deeply rooted kinship established through their social and cultural connections known as ‘gandong,’ derived from Ambonese, which means kandung (siblings) in Indonesian. Gandong, as the focal point and wellspring of life (Tutuarima et al., 2017), resides between two or more villages based on history, blood, or hereditary relations by depicting a pleasant and harmonious relationship pattern. Owing to their socio-cultural affinity, these two communities engage in various sacred endeavors, including constructing Baileu or Baileo (A Moluccan traditional house used for gatherings, meetings, ceremonies, and rituals) and mosques. The interaction between the two communities subsequently facilitated the exchange of lexical items and grammar across their respective languages.

The percentage of cognates shared between Wemale compared to the Alune, Luhu, and Lisabata languages is less than 50%. The geographical distribution of the Wemale group in the eastern and southeastern regions and the Alune group in the northwest part of Western Seram (Collins, 1982, 2018; Knaap, 1993) highlights that geographical distance encourages language and dialect variation. Migrations have superseded their replacement. Through time and space, when speakers of the same group are geographically apart, they tend to develop linguistic distinctions. This phenomenon is referred to as ‘contagious diffusion’, according to Wolfram and Schilling (2016, p. 144).

Due to Luhu and Lisabata’s strong connection with the Alune, Wemale exhibits the lowest degree of kinship among all the languages based on lexicostatistical calculations. As two descendants of the Three Water Trunks, Alune and Wemale ethnic groups have identified as two distinct tribes, have their own language, and possess several distinctive cultural features (Knaap, 1993, p. 254). According to the oral tradition, Alifuru originates from the Western Seram and can be classified as Alune-Melayu Polynesia and Wemale-Melanesia (Bartels, 2017, pp. 40–41). This distinction makes Wemale less reliant on the Alune, Lisabata, and Luhu languages.

4.3. Behind the evolutionary history: the determination of the Alune, Lisabata, Luhu, and Wemale language genealogy through glottochronological calculations

Determining the genealogy of the languages in the Western Seram region does not stop at the language grouping stage. Since lexicostatistics provides a minimal view of language history (Nurse, 1997, p. 359), another interesting aspect that needs to be considered is the separation time, known as the glottochronology aspect for reconstructing ancestral language by identifying their possible time separation. While lexicostatistics establish a relative chronology of language development, glottochronology is a subset of lexicostatistics used to develop absolute dates for those developments over a millennium (Nurse, 1997, p. 366). The development of Nunusaku languages (consisting of the Alune, Lisabata, Luhu, and Wemale) based on their time separation estimation can be seen in Figure 5..

Figure 4. Cladistic (tree-based) representation of the Western Seram language group’s diversification based on a lexicostatistical approach.

Figure 5. The development of Nunusaku language groups based on their time separation estimation.

A language family is a monophyletic unit (i.e. in a sister-group system). It means that all its members derive from a common ancestor and all descendants of that ancestor are included in the family. Through the historical-comparative linguistics approach, as proposed by Swadesh (1952, 1955), language splits can be dated due to the regular decay of words in the basic language lexical item. Therefore, despite many changes in language and lexical items, language proximity can still be investigated by identifying their origin. In this study, the calculations within lexicostatistical and glottochronology supported the position of Alune and Wemale as two early descendants of Nunusaku.

Both Alune and Wemale are ‘insiders’ under the same linguistic tree (Boulan-Smit, 1998, p. 117). As can be observed from Diagram 3, Alune and Wemale separated approximately 1,510 to 1,697 years ago before the separation of other paired languages. This separation can be proven through cultural evidence in the verses of Western Seram’s traditional song, ‘Kapata,’ which was collected during field observation.

Table

Tutu ya hei leta, hei lete o	‘(This is the) verdict from the top, from the top.’
Hei lete Nunusaku o, Nunusaku o	‘From the peak of Nunusaku, Nunusaku.’
Nunusaku samapela o, samapela o	‘Nunusaku has distributed (us), has distributed (us).’
Sama pela waila telu o, waila telu o	‘allotted (us) into Three Trunks of Water, into Three Trunks of Water.’

The Alune and Wemale language separation times provide the basis for identifying other language separation times, including their relationship to the two preceding languages. From glottochronology calculations, it is apparent that Lisabata separated from Alune approximately 1,417 to 1,175 years ago, whereas Luhu’s separation from Alune occurred within the range of 1,466 to 1,214 years ago. Additionally, Lisabata’s separation from Luhu occurred between 1,181 and 950 years ago. It is worth noting that Lisabata diverged from Luhu before its separation from the Alune. These findings then provide clear evidence that Alune, Lisabata, and Luhu share a common ancestral origin. In the meantime, while a potential temporal gap exists in the evolution of the Luhu, Lisabata, and Wemale languages, linguistic kinship identification and lexicostatistic calculations suggest a relatively limited linguistic connection among the three languages. Therefore, it can be concluded that Wemale belongs to a distinct language group or another branch of Nunusaku.

4.4. Phonemic correspondences as qualitative evidence for the Alune, Lisabata, Luhu, and Wemale kinship

The identification of regular sound or phonemic correspondences plays a crucial role in historical-comparative linguistics as the basis for the proof of genetic relationships or the reconstruction of protoforms (Antilla, 1989; Campbell & Poser, 2008; List, 2019). Based on 200 Swadesh’s wordlist lexical items, several forms of correspondence were discovered between Alune-Wemale and Lisabata-Luhu and between Alune-Lisabata-Luhu and Wemale. Several forms of correspondence revealed similarities between Alune, Luhu, and Lisabata. In the meantime, they also revealed the exclusion of three languages from Wemale. The following rules were discovered as evidence for the existing phonemic correspondences.

4.4.1. /ø/∼/y/#_

The rule /ø/∼/y/#_ appeared frequently in four existing languages. In this case, Wemale displayed innovation by inserting prothetic /y/ [+semi-vowel, +lamino-palatal, +voiced] before nouns and pronouns, which would otherwise display the initial vowel /a/. Missing sound (zero phonemes) represented by ‘Ø’ symbol occurs in the initial position of the Alune, Lisabata, and Luhu construction. The whole correspondence pattern within /ø/∼/y/#_ rule can be seen in Table 6.

Table 6. Phonemic correspondences within /ø/∼/y/#_ rule.

No	Gloss	Alune	Lisabata	Luhu	Wemale	Phoneme correspondence
1.	I	au	a?u	a?unɛ	yau	/ø/∼/y/#_
2.	dog	asu	ana	aninɛ	yasu
3.	fire	au^wɛ	anin	asu	yahu
4.	you	alɛ	asu	au	yale
5.	we	ami (itɛ)	ane	alɛnɛ	yami

4.4.2. /b/∼/h/#_, /b/∼/h/V_V#

/b/∼/h/ rule was consistently found in three languages by excluding Wemale from this rule. The consonant /b/ [+stop, +labiodental, +voiced] that occurred in Alune was consistently reflected as consonant /h/ [+fricative, +laryngeal, +voiced] in the initial position of Lisabata and Luhu. However, in rare cases, /b/∼/h/V_V# rule was applied in Alune, Lisabata, and Luhu. Consonant /b/ [+stop, +labiodental, +voiced] in the inter-vowel position of the Alune was shifted to consonant /h/ [+fricative, +laryngeal, +voiced] in the inter-vowel position of Lisabata and /p/ in the inter-vowel position of Luhu since /b/ and /p/ were located in the same area of articulation. The overall correspondence pattern within /b/∼/h/#_ and /b/∼/h/V_V# rules can be seen in Table 7.

Table 7. Phonemic correspondences within /b/∼/h/#_, /b/∼/h/V_V# rules.

No	Gloss	Alune	Lisabata	Luhu	Phoneme correspondence
1.	stone	batu	hatu	hatu	/b/∼/ h /#_
2.	fruit	bu^wai	huwa’i	hu^wa
3.	moon	bulanɛ	hulan	hulane
4.	kill	bunu	hunu	hunu
5.	wife	bina	mahina	mahinaɛ
6.	fly	ribu	rihu	ripu	/b/∼/h/V_V#

Lisabata and Luhu groups tend to show a common language development. This finding aligns with the percentage of kinship between the two unified languages, which reaches more than 60%. Lisabata and Luhu have the same development path in social life due to their migration history from mountainous to coastal areas. Interestingly, even though they have formed different cultural identities and received much religious influence, they still remember and realize that they come from Nunusaku, which is then divided into Alune and Wemale. Although most of them still claimed to be Alune, it was acknowledged that there had also been much contact with the Wemale group. The existence of contact and influence between groups with different languages also influences language development. However, elements of similarity can still be traced to the Lisabata and Luhu languages.

4.4.3. /k/∼/ø/#_

/k/∼/ø/#_ rule was found in Alune, Lisabata, Luhu, and Wemale. Consonants /k/ [+stop, +dorso-velar, -voice] in the Alune, Lisabata, and Luhu languages were totally omitted in Wemale in both the initial position of the penultimate syllable (as in glosses ‘stand’, ‘dig’, ‘nip’, ‘eat’, ‘river’, and ‘drink’) and between vowels and consonants (as in gloss ‘know’). The correspondence data within /k/∼/ø/#_ can be seen in Table 8.

Table 8. Phonemic correspondences within /k/∼/ø/#_ rule.

No	Gloss	Alune	Lisabata	Luhu	Wemale	Phoneme correspondence
1.	stand	kɛlɛ	kene	ɛlɛ	ele	/k/∼/ø/#_
2.	dig	kali	kasi	kasi	ali
3.	nip	kɛtɛ	kiki	ki’i	ete
4.	eat	kanɛ	ka:	–	ane
5.	river	kwɛlɛ	–	–	waele
6.	drink	kinu	inu	–	–
7.	know	rɛkwa	rewa	–	–	/k/∼/ø/V_K#

Based on the data, gloss ‘stand’ there is an inconsistency in Luhu’s language. If Luhu language tends to retain the consonant /k/ in the initial position in most data but in the realization of the gloss ‘stand’, this rule does not apply. The realization of ‘stand’ in Luhu is [ele], so it tends to be the same as Wemale. The same is the case with the Lisabata gloss ‘drink’. If other data shows the closeness between Alune, Lisabata, and Luhu, but in this gloss, it was realized to be [inu] with the omission of consonant/k/in the initial position. This naturally happened considering the existence of ethnic mixing and communication across community groups in the period after the split or separation of the Nunusaku tribe. Regional separation would be easy to sort out as it was related to regional boundaries, but not with the use of language. Language is a part of culture inherent in humans; wherever humans are, they know no boundaries, rules, or restrictions on its usage. Thus, mutual borrowing and the influence of other languages can occur. In addition, the amount of different data has happened in a very small percentage.

4.4.4. /ø/∼/h/#_, /h/∼/ø/#_

/ø/∼/h/#_ and /h/∼/ø/#_ rules were found in Alune, Lisabata, and Wemale. Consonants /h/ [+fricative, +laryngeal, +voiced] in the initial penultimate syllable and Wemale’s final position were lost in Alune and Lisabata. The whole correspondence data within /ø/∼/h/#_ and /h/∼/ø/#_ rules can be seen in Table 9.

Table 9. Phonemic correspondences within /ø/∼/h/#_ and /h/∼/ø/#_ rules.

No	Gloss	Alune	Lisabata	Wemale	Phoneme correspondence
1.	how	ila	-	hai	/ø/∼/h/#_
2.	kill	bunu	hunu	punuh
3.	foot	-	ai	hai	/h/∼/ ø /#_
4.	drink	-	inu	hinu
5.	tie	-	eta	heta

Table 10. Phonemic correspondences within /l/∼/n/#, /l/∼/n/_V#, /l/∼/n/V_V#_ rules.

No	Gloss	Alune	Lisabata	Wemale	Phoneme correspondence
1.	two	luwa	nua	luwa	/l/∼/n/#_
2.	light	lanitɛ	nanite	lante
3.	five	lima	nima	lima
4.	sun	lematai	niamatai	liyamatai
5.	land	tapɛlɛ	ta:pen	-	/l/∼/n/_V#
6.	laugh	mali	mani	-	/l/∼/n/V_V#
7.	three	tɛlu	tenu	tɛlu
8.	bone	lulini	nuni	luli
9.	wind	-	anin	yaline
10.	you	ale	ane	yale
11.	rain	ulanɛ	unan	ulane
12.	head	ulu bu^wa	unu	ulu	/l/∼/n/V_

There were inconsistencies in the realization of gloss ‘kill’ in Lisabata. In glosses ‘how’, ‘foot’, ‘drink’, and ‘tie’ consonants/h/were realized in Wemale, while in gloss ‘kill’, consonant /h/ was discovered in Lisabata aligned with the realization of prefix/p/in Wemale with /h/ in the final position. These findings prove discrepancies with the general rules of phonemic correspondence in some of the realization data.

4.4.5. /l/∼/n/#_, /l/∼/n/_V#, /l/∼/n/V_V#

/l/∼/n/#, /l/∼/n/_V#, /l/∼/n/V_V#_ rules could be found in Alune, Lisabata, and Wemale. Consonants /l/ [+Apiko-alveolar, +voiced] in Alune and Wemale turned into consonants/n/[+nasal, +apico-alveolar, +voiced] in Lisabata at the initial of the penultimate syllable, end of the penultimate syllable, final, and between vowels positions. The whole correspondence data within /ø/∼/h/#_ and /h/∼/ø/#_ rules actually can be seen in Table 10.

Based on the observations and analysis of four languages in the Western Seram region, it shows that there is a genetic relationship. These findings were in accordance with Collins’s Collins (1982, p. 88) state that all the indigenous languages of Western Seram appear to be descendants of a single ancestral language, labelled elsewhere as ‘Nunusaku’. The quantitative evidence classified in the previous discussion is in line with the qualitative evidence in the form of phonemic correspondences between the four languages, aligned with the social relations of society’s historical and cultural perspectives. Many mythological narratives in the Nunusaku tradition recount the origin of the ancestral mountain land, followed by a division and migration toward the coast. The separation of Luhu and Lisabata towards the coast is supported by some evidence of similarities between the two languages. Additionally, the division of Nunusaku into Alune and Wemale, marking the emergence of two major tribes, revealed familial linguistic connections. The analysis results highlighted various linguistic correspondences that distinguish Alune from Wemale, with some shared linguistic features. Meanwhile, Lisabata and Luhu’s relationship based on community information and historical evidence aligned with the hypothesis on the connection of Lisabata and Luhu to the Alune group.

4.5. Recalling the ancestors: the importance of using historical-comparative linguistics approach to the investigation of tribal kinship in Western Seram

According to history, the Nunusaku people were divided into small groups due to inter-ethnic conflict. Unfortunately, the facts surrounding this conflict appear to be hidden and complicated. Deeper issues concerning the causes of its conflict have become extremely taboo, ‘pamali’ to discuss by the indigenous people and surroundings. In its dynamics, the cultural message, which becomes a mindset for human action living, develops into a hegemonic ideology. As a result, the people of Western Seram will never disclose the conflict as an act of forgetting the past to avoid the possibility of repeating the dark history in the future. Therefore, special approaches are needed to maintain their harmony. As stated by one of the respondents in Malay-Ambonese:

‘katong ini semua satu, sudara gandong, mar karna satu deng lain hal, bikin katong terpisah-pisah, ada Alune, ada Wemale. Tapi katong tu satu, Nunusaku. Katong su sama-sama angka sumpah, seng bole lai ungkit-ungkit soal dolo-dolo, orang tua-tua bilang itu pamali.’ (researchers’ self-documentation, 2022)

‘Basically, we are one, as brotherhood. However, for some reason, Alune and Wemale were separated from us. We are one, Nunusaku. We have taken an oath [that] we are not allowed to bring up issues that happened in the past; the elders call it taboo.’

Recalling their ancestors through language is urgently needed for maintaining, fostering, and promoting unity and solidarity. In this case, the disclosure of language kinship can be offered as a reminder to Alune, Lisabata, Luhu, and Wemale that they are actually interconnected within the language family. Radcliffe and Pequeño (2010) emphasized the importance of ethnicity in fostering group consciousness and collective action. This concept encompasses shared elements such as common ancestry, collective memories, and aspects of group identity based on five key factors: kinship, religion, language, shared territory, and nationality or physical appearance. In this case, language plays a vital role in shaping how conflicts resulting from the confrontation of differing standards, values, and attitude structures unfold and resolve (Holtgraves & Taylor, 2014; Nelde, 1987). Since many conflicts exist whenever incompatible activities occur (Deutsch, 1971; Johnson & Johnson, 2009), language has contributed to the development of a ‘sense of belonging’, as a form of emotional bond or psychological safety.

The study of ethnic conflict continues to benefit from using language as a proxy for culture (Mabry, 2011). This kind of innovative approach, in a more paradoxical way, is desperately needed to address negative issues. If culture can influence language, especially in its lexical item, then language can also influence culture (Antilla, 1989). In the shadow of this concern, by minimizing tension between ‘insiders vs. outsiders’, the evidence about language ancestry can be used for granted as a ‘win-win solution’ for building positive associations as the opportunity or willingness to maintain unity, reconciliation, peace, work, and integrity among language groups.

5. Concluding remarks

This present study revealed the kinship relationship between the Alune, Lisabata, Luhu, and Wemale languages, reviewed using a historical-comparative linguistics approach and inter-ethnic relationships based on shared ancestral history. After analyzing four languages in the discussion section, the following important conclusions can be underlined. First, the four languages of Seram Island had genetic relationships, as demonstrated by family relationships supported by historical and socio-cultural evidence. The division of Nunusaku into Alune and Wemale, which marked the emergence of two major tribes, revealed familial linguistic connections as the basis of separation. Meanwhile, community information and historical evidence supported the hypothesis of Lisabata and Luhu’s relationship with the Alune group. Otherwise, the link between Alune, Lisabata, and Luhu is a little special. Alune and Luhu have a ‘gandong’ social and cultural connection, whereas Alune and Lisabata are related to their ethnic affinities, which are divided into four hierarchies: Inama, which stands for Nuniali; Pati, which stands for Lisabata; Angkota, which stands for Wakolo; and Sailbubui, which stands for Buria.

Second, based on the qualitative evidence, the similarities and consistency of sound changes, such as /ø/∼/y/#_, /b/∼/h/#_, /b/∼/h/V_V#, /k/∼/ø/#_, /h/∼/ø/#_, and /l/∼/n/#_, /l/∼/n/V_V#, become the fact that Alune, Lisabata, Luhu, and Wemale languages were interconnected each others. Thus, as Radcliffe and Pequeño (2010) state that the significance of ethnicity lies in its salience for group consciousness and collective action, which includes common ancestry, memories of a shared past, and aspects of group identity based on five factors, such as kinship, religion, language, shared territory, nationality or physical appearance; these qualitative and quantitative shreds of evidence can be used as a milestone to mitigated potential conflicts and to foster a sense of solidity among ethnic speakers and to increase social bonding between ethnic groups by recalling their ancestors.

However, it was recognized that this study is limited to the investigation of language relationships in four areas in Western Seram. Therefore, expansion of the area and populations is required to fully understand the relationship between regions in Western Seram as well as to get more salient data. Furthermore, this study also opens up vast development opportunities, as the relationship between language and identity covers many issues. Language does not exist in a vacuum; somewhat, it is shaped by the historical, cultural, and political context in which it emerges. As the impact, deeper exploration through language reconstruction appears to be critical to obtaining the proto-languages of the languages so that the etymon form can be used as a comparison.

Acknowledgements

We would like to express our deepest gratitude to the Language Office of Moluccas, Indonesia, as well as the Raja, Saniri, and indigenous Alune, Lisabata, Luhu, and Wemale, for their significant role in providing linguistic and cultural insights, which enriched our study. We also give our thanks to the anonymous reviewers who dedicated their time and expertise to assessing our manuscript critically. Their constructive feedback and insights have substantially improved the outcome.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

We would like to express our sincere gratitude to The Indonesia Endowment Funds for Education (LPDP). This study was funded by the Directorate of Research Universitas Gadjah Mada (Dit.lit UGM), Indonesia, under the scheme of Rekognisi Tugas Akhir (Final Project Recognition) program with the Grant Number 5075/UN1.P.II/Dit-Lit/PT.01.01/2023.

Notes on contributors

Hendrokumoro

Hendrokumoro is a senior lecturer in the Faculty of Cultural Sciences at Universitas Gadjah Mada (UGM), Indonesia. His areas of expertise lie in language, communication, and culture; as well as historical-comparative linguistics (HCL), dialectology, Javanese literature, cognitive linguistics, and general linguistics.

Faradika Darman

Faradika Darman is a final-year student in the Master in Linguistics Study Program at the Faculty of Cultural Sciences, Universitas Gadjah Mada,as well as a civil servant in Language Office of Moluccas, Indonesia. Her research interests are descriptive linguistics, as well as Austronesian and endangered languages.

Neni Nuraeni

Neni Nuraeni is a final-year student in the Master in Linguistics Study Program at the Faculty of Cultural Sciences, Universitas Gadjah Mada, Indonesia. She focuses on Austronesian language and descriptive linguistics studies.

Nadia Khumairo Ma’shumah

Nadia Khumairo Ma’shumah graduated with a Master of Arts (M.A.) in Linguistics at the Faculty of Cultural Sciences, Universitas Gadjah Mada (UGM), Indonesia, in 2023. As a junior researcher, her interests lie in the contestation of language, identity, and culture; literary translation; cognitive linguistics, systemic functional linguistics (SFL); as well as critical discourse analysis (CDA).