I. Opening statement
On March 8, 2002, while living in Osaka, Japan, I streamed an episode of National Public Radio Science Friday hosted by Ira Flatow that featured Steven Bird, Jerold Edmondson, and Lawrence Kaplan in a discussion on the topic of language endangerment. It was a topic that I had been tangentially aware of but never previously given much thought to. As a child and young adult I had a deep interest in language and had dabbled in learning several—usually learning more about languages than developing any reasonable degree of proficiency—but after many years of struggle and a year abroad at the Rheinische Friedrich-Wilhelms-Universität Bonn in Germany, I did eventually attain a respectable degree of German language proficiency and having lived in Japan for nearly a year at that time, I had begun to develop basic Japanese skills as well. Although my successes were matched by a deep frustration about seemingly insurmountable obstacles I sensed in my post critical period language learning attempt. I recounted that as a small child living in a Zen Buddhist monastery outside of Woodstock New York, I had already known quite a bit of Japanese—the exact amount I am uncertain, as it was a process that came to me naturally from my surroundings. And while I have a memory of once being ushered into the prayer room and asked to lead the adults in a mantra, this was no doubt done on a whim to satisfy curiosity or amusement from adults who noticed me as a mischievous little distraction to worship while playing in the surrounding hallway and attic spaces during their prayer activities, and I did not likely understand much of what I was reciting (if it was even Japanese at all).
The mantra aside, I was able to speak a fair bit of Japanese, and those decades later I could vividly recall the feeling of having previously understood and spoken a language that I had effortlessly and unintentionally absorbed, yet the singular word I could still produce from my childhood was the ‘Buddhist name’ I had chosen for myself when the elder priest, Daido, shaved my head, dressed me in a robe, and declared me “the Zen Mountain Monastery’s newest and youngest monk.” Incidentally, in a characteristic display of humility, I chose for my name a Sanskrit word maha—a word that traces back to Proto-Indo-European *meĝ- meaning ‘big, much, great, major, mega-‘. To put it a different way, the experience of effortlessly learning a foreign language when I was five was followed by effortlessly forgetting it all the following year when I moved away, and it afforded my adult self no advantage when I started trying to intentionally learn Japanese. Language learning had become a struggle and I realized that no matter how I might try, I would likely never reach a level approaching native proficiency.
The radio conversation evoked a powerful emotional response, making me consider the loneliness of being the last speaker of one’s own mother tongue. No matter how I might try, I would likely never reach a level approaching native proficiency that could convey my deepest thoughts and feelings in a way close to representing my actual human experience, and realizing this overwhelmed me with a sense of loneliness at the world’s irrevocable loss when a language ceases to exist. In the years following, I often thought about the ongoing issues of language loss and wondered how I could help address the problem. I committed myself to pursuing linguistics, because I did not want to risk unconsciously exacerbating the problem. I set a somewhat arbitrary goal of earning a footnote or citation to mark the moment I would become linguistically ‘carbon neutral’ and I decided that anything beyond that point would constitute a net social contribution.
Although I wanted badly to help, I knew I was not likely to become a revolutionary theoretical linguist; I was long out of college and did not have a background in the field, and to be honest, the back and forth game of theoretical academics has never brought me satisfaction nor a sense of concrete contribution. I thought academia was probably not the most suitable path for me, but as I studied the topic of language loss, I kept hearing a common refrain from the field’s preeminent linguists: languages are dying faster than they can be studied, and there are simply not enough linguists to gather even the most basic information about languages before they are gone. This was often followed by an impassioned plea for people to become linguists and work on an undocumented language. That was a contribution I could see myself making, as it didn’t require expanding the cutting edge of linguistic theory.
Where there is terra incognita, cartographers must first map the landscape before the geologist can know where to find the canyons and arêtes she studies, before the prospector can make an educated guess about where to sink his shovel, and before the ichthyologist can know where to find lakes and rivers. While the cartographer necessarily has a much shallower base of knowledge about each of those topics, and her work is arguably less glamorous than other areas, her initial mapping is vital, as those areas of research fundamentally depend on her groundwork. Likewise, theoretical and experimental linguists of today must rely on rudimentary wordlists and analyses collected by previous fieldworkers in order to expand the base of knowledge. This dissertation is that contribution to linguistic knowledge: a modest but broad cultural (and sometimes physical) mapping of the landscape regarding Sula: its language, land, and people.
My first introduction to the Sula language came during volunteer work with the Language Documentation Training Center (LDTC) at the University of Hawaiʻi at Mānoa. There I was partnered with Erwin Gay, a man originally from the village of Pohea on Sanana whose early life experiences took him to Ternate and eventually led to a university study abroad program in Hawaiʻi. Erwin joined the LDTC to share a part of his language and culture with the world, and as I began to work with him and search for information and previous research on the language, it became clear that the Sula language was underrepresented in the academic literature and few of the answers to my questions were available. As someone who does not easily accept his curiosity going unanswered, my initial introduction to the language and culture of Sula ignited an ever-lengthening chain of questions and a (sometimes obsessive) decade-long labor of passion which culminated in this dissertation: a summary of findings on the language and people of the Sula Archipelago. I hope that the information herein will assist with aid and development efforts in the region, facilitate language maintenance and revitalization efforts, kindle future in-depth linguistic and anthropological research, and provide data that could one day help to refine linguistic theories and improve our science’s understanding of language mechanics and its interface with the human mind.
This dissertation represents the culmination of my effort to delve deeper into a little-known language in an effort to document and provide information about at least one more world tongue before it possibly becomes extinct. When I embarked on this study, so little was known about the Sula language in the academic literature that even the most basic question could not be answered: is Sula a singular language, and if so, is it endangered? Fieldwork for this dissertation was conducted in the Maluku region of Indonesia, primarily on the islands of Sanana and Mangoli within the Sula Archipelago and also among diaspora communities on Ternate, Surabaya, and Halmahera. Work with one of my most valuable collaborators, Ida Ryberg (née Tabona Umage), was conducted in Gothenburg, Sweden and thereafter via ongoing digital correspondence.
Regardless whether Sula turned out to be endangered, I reasoned that my initial work on Sula would help document a dying language, because even if Sula turned out to be healthy, my work would let other linguists know to divert their energy to other languages in need of more pressing attention. Essentially, even if my work did not itself amount to direct documentation of an endangered or dying language, it would at a minimum help to do so indirectly. So while this dissertation does not attempt to propose any revolutionary theories on human language, it does identify areas where the language challenges accepted linguistic dogma, and in a broad sense, it answers the questions: Is Sula a single language or multiple languages? Where is Sula spoken and by whom? What is its basic lexicon? How are the sub-dialects (or languages) historically related to one another? How is Sula’s basic grammar structured? (with regard to the core areas of linguistics). Lastly, What is life like for Sula speakers? (concerning geographic, demographic, psychosocial factors, ethnographic, and infrastructural information).
Upon commencement of research it was unclear what the most interesting or theoretically poignant aspects of Sula life and language would be, but years of work, significant personal investment, and the thoughtful guidance of my mentors, advisors, and collaborators, has led to the identification of several such areas. This directs experimental and documentary linguists toward important areas for subsequent in-depth studies so they can hit the ground running and tease out valuable answers. It also provides easily accessible, hard data for academics who themselves operate on the cutting edges to tests their theories and to identify further poignant topics that slipped past me.
The classical and modern giants in the field of linguistics established a firm footing for me to operate on, and they provided time-tested conventions and methods for those in my shoes to utilize. So although my work represents only the tip of a much bigger iceberg in understanding Sula (or perhaps ‘pumice island’ would be a more regionally appropriate metaphor), and while this document certainly contains mistakes and analytical flaws, I can state confidently that the information herein is sound and useful, and that it will lead to meaningful contributions to our field in the years to come.
II. Dissertation context
Globalization has begun to connect Sula to the outside world, and the language community is on the precipice of a profound social restructuring. Sula is endangered by policies that bolster Indonesian Malay to the detriment of indigenous tongues, by the introduction of digital technologies that deliver dominant-language content to previously private domains of communication, and by a depressed local economy that requires many in the community to seek jobs elsewhere. This dissertation represents nearly a decade of doctoral work geared at helping to curtail threats to language vitality. It includes a Sula language grammar, a Sula–Indonesian–English dictionary, documentation of traditional environmental knowledge, and annotated textual examples of audio-visual materials that capture Sula in culturally relevant discourse domains—especially domains at risk of disappearing. Herein you will learn about the inner mechanics of Sula and its external pressures alike.
The Sula language of Eastern Indonesia is an Austronesian language of the Central-Eastern Malayo-Polynesian sub-family (Eberhard et al. 2020) that is spoken (to varying degrees) by likely upwards of 40,000 people on the islands of Sanana and Mongoli and (to a lesser degree) Taliabu, Buru, Bacan, Ternate, and in Surubaya.
The language has twenty consonants and five vowels. It is a subject-verb-object (SVO) language with position indicated by a prepositional locative morpheme working in concert with meaning-bearing postpositional phrases. The genitive is possessor initial, and modifiers follow the noun—that is, mon nap nahu (you hair long) ‘your hair is long’. This is somewhat uncommon typologically, as languages with post-nominal modifiers tend to also have post-nominal possessors (e.g. Comrie 1989). Most Sula words are one or two syllables; however, word stems of up to four syllables are not uncommon, and can be much longer in compound words and words with reduplicated modifier suffixes. Canonical syllables are shaped (C)V in the Mangon dialects of Sula and (C)V(C) in the Sanana dialects—that is, the majority of Proto–Sula words were bisyllabic, but the Sanana dialect has deleted many final vowels that are still present in Mangon. This has resulted in many more monosyllabic words in the Sanana dialects and many syllable codas that are not present in Mangon dialects. Grammatical words tend to be monosyllabic, and disyllabic words carry stress on the penult if the final syllable is light but on the final syllable, if it is heavy (in a moraic analysis, stress could be said to be present on the syllable containing the penultimate mora).
Sula has a rich and complicated pronominal system; pronouns are frequently dropped, and they are indexed on verbs via a set of agreement markers that are also used (by many if not all speakers) to indicate the progressive aspect and reference switching. As in the Buru language (Grimes 1992), deictics are used to indicate definiteness and reference tracking. Sula has a system of split alignment and animacy (at least among many speakers): first-person subjects follow an Active–Stative classification in which transitive verbs and agentive intransitive verbs are marked for subject agreement. Non-first-person, human subjects are marked on verbs along ergative–absolutive lines, and verbs are not marked to agree with non-human subjects. Sula does not have grammatical tense; however, there is a system of post-verbal aspect markers. Sula does not seem to have a very rich system of relativization or clause embedding, and where it is observed, it is unclear if it is native or calqued from Malay.
II.I. Research Overview
This dissertation includes #### thousand Sula lexical items, a grammatical description of the language, a description of the primary dialects, and a historical reconstruction of the proto dialect. The work supports the Sula community’s efforts to preserve its language and heritage, and it contributes to the body of knowledge about a geographically large area that remains nearly unknown to the linguistic and anthropological communities.
Research is based on data collected during fieldwork that began in 2010. Language vitality is determined using a twenty-factor augmented version of UNESCO’s factors for determining language endangerment (Brenzinger et al. 2003). My data include grammar and vocabulary elicitations; interviews; and recordings that capture Sula in culturally relevant discourse domains—especially domains at risk of disappearing, such as: casual conversations, public events, demonstrations of cultural practices, and performances of songs and stories that are connected to physical locations in the archipelago (Bloyd 2010–2019). The words of the Sula language are collected through targeted vocabulary elicitation and examination of conversation samples, and the entries are compiled into an online and downloadable Sula–Indonesian–English dictionary.
Linguists commonly note that work on Sula is limited by a deficiency of descriptive data (e.g., Blust 1981, Collins 1981, 1983, Esser 1938, Grimes 1992). The dissertation’s lexical documentation will facilitate historical analysis, the grammatical sketch will enable typological and syntactic research into the boundaries of what is possible in human language, and the dialect descriptions will make possible sociolinguistic variation research. These are lines of inquiry that the paucity of data has rendered thus far impossible to conduct.
The research benefits the Sula community, some of whom have become increasingly concerned about the trend of language loss and hope to preserve their language heritage: the language vitality component provides information that will assist planning efforts, and documentation materials will assist language maintenance and revitalization programs and aid with the production of pedagogical materials. This work has already begun to mitigate absolute language loss from the standpoint of documentation, as I have trained three Sula speakers in methods of documentation, and I am in the process of training several more. The Jere Feu project for local ecological knowledge in the modern era that I began with my colleague, Adita Agoes, works to preserve local regional languages through the framework of Jere, the region’s traditional knowledge and conservation system. This project has over fifteen partners including community elders, government officials, educators, and it even boasts support from the prestigious sultanate of Tidore and the late sultan of Ternate.
III. Sociopolitical context during the time of research
I made my first fieldwork trip to Sula in Summer of 2010. It was a time of transition and optimism. Barak Obama had been inaugurated in the United States a year before, and he not only ended a perceived period of hostility toward Muslims worldwide, he had himself lived in Indonesia during a period of his childhood, has Indonesian family, and could at least pay lip service to the the nation’s official language. I enjoyed a good deal of favor by association as I was a researcher from the University of Hawaiʻi—Obama’s home state—and a student affiliate of the East West Center academic institution where Obama’s parents had met, and I lived on the third floor of the Hale Mānoa building in a room where (according to plausible urban legend) Obama’s parents lived and he was himself said to be conceived. I must admit that I used that tenuous connection to grease wheels and get myself out of a bind on more than a few occasions.
A characteristic feature of my first summer in Sula was the 2010 World Cup—an event that evoked a strong sense of positivity, hope, and international unity among Moluccans. Indonesia is not known to be a football powerhouse on the world stage, and they were by no means contenders for the competition, but Sanana residents were as ferociously dedicated as any sports fans I have ever observed. Each neighborhood chose a country to support—nations that were foreign in the purest sense, places where few if any of Sanana’s residents had ever set foot and whose languages and cultures were largely unknown, yet places many of Sanana’s young men would no doubt have taken up arms and given their lives for.
Several times a day throughout the duration of the competition, large makeshift parades rallied down the main strip supporting the nations set to compete. Participants would wave whatever flags and paraphernalia they could make or acquire, and blast out sound from whatever noise-making devices they could render. During game times, crowds gathered around public outdoor televisions powered by community generators to watch, celebrate, and cheer their adopted nations on.
During the course of my research, the Islamic holy month of Ramadan spanned the (northern hemisphere’s) summer months, when my academic and work schedules permitted most of my field work, and the observance had a significant influence on my research. On the positive side, the observance encourages a general feeling of positivity and love for one’s brethren, but on the flip side, the fasting requirement greatly impacted my consultants’ ability to focus on work. During the Holy Month, Muslims are encouraged to abstain from eating or drinking from sunup to sundown. The Sula Archipelago is stiflingly humid during that time of year, however, and the resulting dehydration can quickly cause lethargy. Additionally, the Sula population are intensely interested in others’ comings and goings, and this led to a number of uncomfortable encounters even while working with consultants who were less severely affected by dehydration: frequently while walking to research appointments, bystanders gathered in public sitting areas would call out to my consultants and assistants with a passive aggressive (but unmistakably judgmental) accusation that they were not observing the fast.
The situation confused me the first time it happened and I asked my assistant whether she knew the bystanders and how they knew whether or not she was fasting. She said that she did not know them but explained to me the logic behind their statement: since people who fast from water during Ramadan tend to become fatigued in the midday heat, most people remain sedentary indoors unless it is absolutely necessary to go out. Because she was out, she said, their assumption was that she must not be observing the fast. This irritated her greatly because she was indeed observing the fast and, as she put it, even if she wasn’t, they did not know whether she was menstruating (one of several reasons a person is permitted to skip fasting). She aggressively scolded the men for their intrusion into her business as we walked on.
My male assistants also faced particularly aggressive ‘nosiness’ regarding comings and goings during Ramadan, but it was more common and far more aggressive when targeting my female assistants. The encounters greatly upset one assistant in particular who hinted that there were additional implications of their accusation, but she did not spell them out, so I can only speculate as to what the implications might have been. When such an encounter occasionally began to grow intense, I learned that I could diffuse the situation with a gentle reminder to the parties involved that it was the Holy Month and we are meant to reject any negative thoughts that enter the mind.
Although my research began during a period of stability and optimism, it is important to note that it was a young stability coming out of tumultuous period marked by religious conflict and political reorganization. I took my first trip to Indonesia in the late 1990’s while on a semester break from college. During that trip, I made a series of wrong decisions that resulted in my being lost somewhere in a seemingly evacuated district of Jakarta. There was not a soul to be seen other than groups of soldiers who were perched atop military vehicles at each intersection, intently scowling at me and clenching their rifles as I aimlessly wandered by. I grew nervous the longer I was lost, and then in the distance I saw a large group of protesters waving black signs and banners moving rapidly toward me. All of the US State Department travel advisories and international news stories about Islamic revolutionaries and rioters had me terrified and unsure whether to fall in with the protesters angrily approaching from my front or try to get past the heavily armed soldiers behind and to the sides of me. A healthy aversion toward approaching scowling people clutching weapons made me take my chances with the protesters. When they reached me, rather than attacking me or hurling projectiles or insults, they rushed toward me with smiling faces asking me to take their photographs. To my surprise, several of them spoke excellent English, and they were eager to help me find my way. I would later learn that such student protests managed to topple a dictatorship and usher in a new era of democracy. Unfortunately though, upsetting the balance of power also led to violent regional flare-ups around the country. One of these flare-ups was a period of religious warfare in Maluku that continued until early 2002 when the Malino II Accord was signed, officially bringing an end to the conflict. Heavily damaged buildings in parts of Maluku bear witness to the strife to this day.
Although the violence was severe in several parts of Maluku, it is unclear to what degree the islands of Sula were involved. News reports from the time mention two Christians on Sanana being burned to death (Mardai 1999), but few details are given about the incident and the sourcing is dubious. Rumors spread with abandon during and following the conflict, and for several years foreign visitors were allowed by permit only, hence international reporters had few means to gather information firsthand. Producing an accurate accounting of the scope of the violence remains no light task. I failed to find firsthand witnesses to the burnings or other related instances of sectarian violence in Sula, but it is safe to say that some degree of conflict happened there, though the situation was not as intense or widespread as in other parts of the region.
IV. Research Justification
IV.I. Research Justification. Broad
This dissertation focuses on a language community nestled within the Austronesian–Papuan contact-region of Eastern Indonesia. It addresses the questions: What constitutes the Sula language? How is Sula distinct from other languages? How has the Sula language evolved over time to reflect the contexts in which it emerged? and what is Sula’s social and physical environment? The questions are addressed by sketching Sula’s grammar (phonology, morphology, lexical categories, and syntax) and by looking at language typology, dialect variation, and identifying contextual factors that affect linguistic viability, including speaker numbers, intergenerational language transmission, cultural and political pressures, regional economy, community attitudes about the language, and the effects of what Grenoble (2011) identifies as primary factors associated with language shift: urbanization, globalization, and social and cultural dislocation.
This work will benefit the Sula community in a number of ways. The language vitality assessment provides answers about the language community that are badly needed for planning efforts, and the documentation materials within will be of use for language maintenance and revitalization programs and aid with the production of pedagogical materials. The academic community will also benefit from my work, because, as Sula is an undocumented language, linguists have few sources of information.
My research is based on data collected during ongoing fieldwork that began in 2010. The data include elicited vocabulary and grammatical information; recordings of free conversation, songs, stories, interviews, public events, demonstrations of cultural practices with native-language explanations, and performances of songs and stories connected to physical locations in the Archipelago. The words of the Sula language were collected through targeted vocabulary elicitation and examination of conversation samples, and entries have been compiled into a sizable dictionary in Chapter 4 of the dissertation.
IV.II. Research Justification. Academic community
The results of my research represent the largest Sula documentation to date. Few publications discuss the Sula language, and fewer still contain any primary lexical data. Holton 1996 lists eight entries that mention Sula, and of them, only four contain primary lexical data (one of which remains unpublished). These entries are: Collins (1976, 1981,) Holle (c. 1900) via Stokhof (1980), and Wallace (1869). Comparative linguists have noted that work on Sula is limited by the deficiency of descriptive data (e.g., Blust 1981, Collins 1981, 1983, Esser 1938, Grimes 1992). Two neighboring languages have been partially described: Fortgens 1921 sketches Soboyo grammar on the neighboring island of Taliabu; and Devin (1989) and Grimes (1992) have described the main indigenous language of Buru island immediately to the south, where there is a sizable community of Sula diaspora in the village of Namlea.
Sula’s precise genetic subgrouping remains unsettled. Based on the material available at the time, Blust (1981) and Collins (1981) argued for an Austronesian subgroup that includes Buru, Sula, and Taliabu. Donohue and Grimes (2008) challenge the Central Eastern subgrouping favored by Blust, and the Blust follow-up (2009) defends the grouping. No new data source was cited either as basis for the challenge or the follow-up. Unsettled matters in Austronesian linguistics such as this underscore the importance of accurate and in-depth descriptive work of the kind presented in this dissertation.
One chapter of my dissertation includes a grammatical sketch that will provide new data to typologists and syntacticians surveying the boundaries of what is possible and what is unattested in human language. Another chapter provides a dialect map of Sula and a description of the phonological characteristics that differentiate each dialect and a reconstruction of Sula’s proto ancestor. The dialect map aids future sociolinguistic research on variation in the Sula language, and the reconstructed Proto–Sula forms aid historical linguists in comparing Sula with its neighbors to refine or verify its proposed placement within the Austronesian language family.
The phonological description includes two phonetically unmotivated sound changes I discovered in Sula that are not known to have ever occurred in another language. The peculiarity of these changes led to their inclusion in a 2018 publication by Robert Blust. These changes pose a problem to the fundamental Neogrammarian hypothesis that sound changes are conditioned only by phonetic factors (e.g. Osthoff and Brugmann 1878, Hock 1991). They also identify a process of intervocalic fortition that is not only synchronically unattested in any other language but one that scholars have asserted to be so unnatural and contrary to universal human language tendencies that it can be assumed not to exist (e.g. Beguš 2015). This documented alternation lends synchronic corroboration to a contested instance of historical devoicing that Robert Blust discovered to have occurred in Kiput and Berawan (Blust 2002, 2005, 2013).
Although Sula belongs to the Austronesian language family, many of its words have unknown origins. Much of this vocabulary is gathered in the dictionary chapter and made available for comparison to other languages. Such comparisons could reveal ancient connections between civilizations or an updated understanding of human migration patterns—or at least help answer some of the persistent questions about how and when humans settled in the region.
The islands’ remoteness creates a partial buffer to outside forces, and documentation is still an attainable goal. But the buffer is rapidly eroding, and Sula’s uniqueness and unusual historical and geographical circumstance demand it be documented in its current context, as globalization has begun to connect Sula to the outside world, and the language community is on the brink of a profound social restructuring. My research is valuable to the Sula community, and it is a vital addition to the academic literature—both because it concerns an undocumented language and because it challenges some long-held assumptions about what is and what is not possible in human language. Without adequate research, the language could cease to be spoken before it is documented. If the language contains as yet undiscovered cognate words, grammatacal patterns, or even folk histories, its disappearance could mean the loss of part of our prehistorical record and, potentially, also loss of some of the keys necessary for unlocking the mechanics of human language and cognition.
V. Chapter overviews
Chapter 1 of this dissertation provides basic sociodemographic information about Sula and continues on to investigate the state of technological infrastructure within the archipelago: specifically the degree of access to Internet-capable digital devices and device hardware/software typology. The information addresses the types of digital services provided in different areas, cell tower locations, areal bandwidth capabilities, and determines the actual and adjusted cost of access to digital communication. These areas of investigation assess the reach of digital communication into the Sula speaking community. Chapter 2 describes the primary branches of the Sula language from a comparative linguistic perspective and it explores the creation of a newer blended dialect. It also describes a controlled series of parallel documentation efforts across many villages in the archipelago that is balanced for age, gender, and tribal affiliation. This documentation will be provided to the academic community and serve as a basis for future projects on dialects and social variation. Chapter 3 provides a grammatical sketch of the language that covers the most fundamental topics within the core areas of linguistics, and Chapter 4 provides the largest known dictionary of the Sula language, containing ### thousand entries, mostly trilingual. The appendix contains supporting materials and resources for others wishing to study the language. Additionally, there is a strong digital component to this dissertation; its most useful form will be its online version that hyperlinks between sections and to hours of supporting audiovisual content.