1. Introduction

Since 1965, when I first began to record and attempt to understand the ritual poetry of the Rotenese, the study of semantic parallelism has been a major focus of my research. My chief concern has been with the semantics of canonical parallelism. In my view, the study of parallelism and particularly the study of canonical parallelism is no minor subject but is, in fact, a matter of central theoretical importance for both anthropology and linguistics. To introduce this collection of papers, let me begin by offering some explanation of the importance I attach to research on this topic.

Parallelism as a focus for research

Parallelism describes the common tendency to resort to the pairing of words and phrases to provide emphasis, authority or significance to an expression of ideas. It is a common, frequently used rhetorical device in many forms of elevated speech. It is also a recurrent feature of poetic discourse—what the linguist Roman Jakobson has called the ‘poetic artifice…of recurrent returns’, or, in a similar vein, what the poet Gerald Manley Hopkins referred to as ‘the repeated figure’ in poetry. Examples of such parallelism are numerous and can be encountered daily. Winston Churchill’s speech to the Commons on 4 June 1940 is a ringing example of the use of parallelism. I have intentionally arranged this passage to highlight its rhetorical pairings:

We shall not flag nor fail.

We shall go on to the end.

We shall fight in France

and on the seas and oceans;

We shall fight with growing confidence and growing strength

in the air…

We shall fight on beaches,

landing grounds,

in fields,

in streets and on the hills.

Canonical parallelism is something different: a linguistic phenomenon that is more specific and more circumscribed but for that reason, theoretically more engaging. Canonical parallelism draws on the tendency involved in the common use of recurrent, parallel phrasing but it defines what terms and phrases may form pairs. This culturally defined, more obligatory aspect of parallelism constitutes what I refer to as ‘canonical parallelism’.

In 1753 Robert Lowth adopted the term ‘parallelism’ (or in Latin, parallelismus membrorum) to describe the linguistic phenomenon that he encountered in his research on Hebrew poetry: ‘a certain equality, resemblance, or parallelism between members of each period…as if fitted to each other by a kind of rule or measure.’ Thus the term parallelism was originally used to designate a regulated correspondence of terms and phrases. In time, however, this usage of the term widened and its technical sense became eroded, requiring further specification. Roman Jakobson introduced the idea of a ‘canonic parallelism’ and I have appropriated this usage in my research. His designation of this phenomenon, pointing to both its prevalence and its pertinence, is clear:

Those poetic patterns where certain similarities between successive verbal sequences are compulsory or enjoy a high preference appear to be widespread in the languages of the world, and they are particularly gratifying both for the study of poetic language and for linguistic analysis in general. Such traditional types of canonic parallelism offer us an insight into the various forms of relationship among different aspects of language. (Jakobson 1966:399)

The first identification of parallelism as a distinct means of poetic composition was made in relation to Hebrew poetry, and this Biblical insight gave impetus to a growing recognition of the occurrence of parallelism and its significance in oral poetic compositions throughout the world. Parallelism occurs in widely diverse oral poetic traditions, from Finno-Ugric and Mongolian to Chinese and Vietnamese, as well as in numerous Dravidian, Austronesian and Amerindian languages, with particularly elaborate forms in Mayan languages. Historically, however, the Hebrew traditions of parallelism, which have now been linked to similar traditions in Canaanite poetry, were the starting point for the study of parallelism. These traditions form part of a more ancient tradition that extends back to beginnings in Sumer when the first known forms of ritual poetry were committed to writing.

Some of the oldest existing examples of canonical parallelism can be found on cuneiform tablets that date to at least 1750 BC. Samuel Noah Kramer spent much of his life deciphering these cuneiform materials and was one of the leading authorities on the use of parallelism in the traditions of the Middle East (Kramer 1979.. In an effort to convey the beauty of these often fragmented texts, he recruited the poet Diana Wolkstein to assist him to translate a set of the ritual texts relating to the goddess Inanna. The following short specimen, an example of this collaboration, describes the beginnings of time before the planting of the sacred Huluppu tree from which both the throne and the marriage bed of Inanna were carved. Its parallelism is evident:

In the first days, in the very first days,

In the first nights, in the very first nights,

In the first years, in the very first years,

In the first days when everything needed was brought into being,

In the first days when everything needed was properly nourished,

When bread was baked in the shrines of the land,

And bread was tasted in the homes of the land,

When heaven had moved away from earth,

And earth had separated from heaven

And the name of man was fixed…1

These lines—and the great corpus of similar compositions from which they come—establish the existence of parallelism as a means of poetic diction that stands at the beginnings of human literature.

In terms of understanding this parallelism, the breakthrough in scholarship came with the recognition that the composition of parallel lines and phrases—Lowth’s correspondence of ‘words to words’—was based on a repertoire of fixed word pairs and, with the discovery at Ras Shamra in 1928 of a cuneiform library of Ugaritic texts, that many of the parallel terms in Hebrew were shared by the Canaanite poetic tradition. This gave rise to an extensive scholarship.2

Although for the most part confined within the defined bounds of its own scholarship, there is hardly a linguistic observation within this Biblical research that does not relate to the wider body of comparative research on traditions of parallelism throughout the world. It is this comparative dimension that is critical.

For this reason, the first paper I have included in this volume is an extended comparative examination of the studies of parallelism and of the major textual sources of parallel poetry. This chapter, ‘Roman Jakobson and the comparative study of parallelism’, which I have substantially updated, was originally written in recognition of the work of the linguist Roman Jakobson, who had a continuing interest in the study of parallelism throughout his lifetime and who wrote profoundly and influentially about parallelism as a linguistic phenomenon.

The Jakobson bequest

I began my own research on parallelism through the personal promptings of Roman Jakobson whom I met soon after I took up a position in the Department of Anthropology at Harvard University in 1969. I had already gathered an extensive corpus of Rotenese ritual chants during fieldwork on the island of Rote in 1965–66. Having completed my doctoral dissertation at Oxford—a thesis concerned mainly with the social organisation of the Rotenese of the domain of Termanu—I was beginning to turn my attention to translating and interpreting some of the chants within this corpus.

The first chant I translated, analysed and published was one of the most beautiful compositions I had recorded. This chant, entitled Dela Koli ma Seko Bunak, was recited by the oldest poet in Termanu, Stefanus Adulanu, who was both teacher and mentor in my study of Rotenese ritual language. My translation and analysis were published in 1971 as ‘Semantic parallelism in Rotinese ritual language’ (Chapter 3 in this volume).3 Citing Jakobson’s ‘Grammatical parallelism and its Russian facet’ as my starting point, I also attempted in this paper to set out the comparative evidence for the use of canonical parallelism, particularly among other Indonesian populations. My principal goal for the paper was to set out as clearly as possible a single, relatively long ritual language text to illustrate the workings of Rotenese parallelism.

At a personal level, the paper allowed me to show Jakobson just what I was working on and to provide some basis for our subsequent discussions. During my time at Harvard, I would meet with him at his invitation and invariably at his home, which was a short walk from my office in William James Hall. For the most part, Jakobson would talk and I would listen. The same was true when we had lunch together and were joined by his wife, Krystyna Pomorska-Jakobson. Jakobson seemed to want to prime me with references to the work of other linguists that would assist me in my own research and would always try to situate his references to particular studies within a personal context. There was no doubt that for him parallelism was a central concern. He referred to the study of parallelism as ‘the double door’ linking linguistics and anthropology.

At the time that I wrote my first paper on Rotenese parallelism, I was one of the supervisors of Gary Gossen’s PhD thesis on Chamula oral traditions, research that was eventually published in 1974 as Chamulas in the World of the Sun. Gossen’s work gave me an initial glimpse into the rich parallel traditions of the Maya and set me off in search of other linguistic traditions of parallelism. These comparative efforts eventuated in my paper ‘Roman Jakobson and the comparative study of parallelism’, which was published in 1977 with 33 other contributions in a tribute volume entitled Roman Jakobson: Echoes of his scholarship. Chapter 2 in the present volume is based on that original paper, but since that paper was written more than 30 years ago, I have felt it necessary to expand upon it and update references to new sources that I have subsequently discovered and other work that has appeared in the continuing study of parallelism.

By the time my tribute to Jakobson appeared, I had moved to The Australian National University to take up a position as Professorial Fellow in the Department of Anthropology in the then Research School of Pacific Studies. Jakobson kept in touch by regularly sending me copies of his latest reprints. In response to my paper, he sent me a reprint, ‘A few remarks on Peirce, pathfinder in the science of language’ (1977), on which he wrote: ‘To James J. Fox with thanks for the convincing parallel to my Parallelism, Roman Jakobson.’

Jakobson’s selection of this particular reprint with its focus on the work of Charles Sanders Peirce was, in my view, intended to send me a message. He was already familiar with the beginnings of my analysis of Rotenese ritual language and my use of the concept of a ‘dyadic set’, which could be specified by the simple notation (a,b), to refer to paired terms. Hence, I took his selection of quotes from Peirce in the reprint—‘[a] dyad consists of two subjects brought into oneness’ or ‘there is an element of twoness in every set’—as explicit encouragement of my efforts of analysis since such notions also underline my understanding of a dyadic set.

Parallelism figured prominently in the framework Jakobson created in his study of language and poetics. His recourse, in Fundamentals of Language (Jakobson and Halle 1956), to the use of opposition to articulate this framework resulted in his formulation of two contrasting aspects of language: a bipolar structure consisting of metaphoric and metonymic poles of discourse. A contiguous sequential arrangement of elements defines the metonymic pole, whereas a selective articulation of elements defines the metaphoric pole. The one relies on contiguity and results in syntactic relations among elements; the other relies on similarity and results in semantic relations among elements. It is precisely in this context that Jakobson made a particularly cogent argument for the significance of parallelism:

In verbal art the interaction of these elements is especially pronounced. Rich material for the study of this relationship is to be found in verse patterns which require a compulsory parallelism between adjacent lines, for example in Biblical poetry or in the West Finnic and, to some extent, the Russian oral traditions. This provides an objective criterion of what in a given speech community acts as a correspondence. (Jakobson and Halle 1956:77)

This understanding of metaphor as semantic selection could be considered a departure point in my research, but from the beginnings of my ‘explorations’ of parallelism, I have been interested in investigating the connectivity of semantic elements. As such, the dyadic set can rarely be considered on its own; its particular terms or elements have invariably to be considered as part of a larger network of relationships.

There are other aspects of canonical parallelism that also require great attention. One of the major goals of my research has been to provide a comprehensive study of one still-living tradition of canonical parallelism.

Arrangement of the volume

The papers that comprise this volume were written over a period of more than 40 years. They chart the development of my understanding of parallelism in general and the Rotenese tradition of parallelism in particular.

This volume is divided into three sections. The first section is made up of eight chapters that deal with general issues of parallelism. These chapters include this introduction, an initial survey of the literature on parallelism, which I wrote as a tribute to Roman Jakobson, and another chapter that attempts to update this earlier survey with an extended examination of recent research on the comparative study of parallelism. The remaining chapters in this section focus on the specific traditions of canonical parallelism among the Rotenese of eastern Indonesia.

Each of these chapters is a distinct exploration of Rotenese parallelism, developing a variety of analytic perspectives and situating the use of this parallelism within its social, linguistic and historical context. In particular, Chapter 4, 5 and 6 were foundational for my understanding of Rotenese parallelism. They were all written in the 1970s as early research efforts. Chapter 4 was intended to present and to analyse a complete Rotenese composition, setting out as clearly as possible the main characteristics of Rotenese ‘ritual language’ as compositions in strict canonical parallelism Chapter 5 was written for a symposium on the ethnography of speaking. In it, I was particularly interested to examine the use of ritual language in relation to ordinary Rotenese discourse and to the use of Malay (and eventually Indonesian), which was adopted as a register for elite communication on Rote from the early eighteenth century. Chapter 6 was written in an attempt to define a symbolic core to the Rotenese ritual canon.

Although each of these papers covers different general aspects of Rotenese parallelism, these paper engage, in varying degrees, in similar modes of analysis, examining, in particular, the relational character of Rotenese parallel semantics and the use of networks as a means of comprehending these relationships. These papers also hint at the use of dialects in ritual language, a subject that became more important in my later work.

By contrast, Chapters 7 and 8 are concerned with other aspects of parallelism. Chapter 7 was written for a symposium on ‘dual organisation’. In the chapter, I am particularly concerned to distinguish parallelism as a linguistic phenomenon from other forms of dual structures that are dependent on the analogical ordering of categories. Parallelism can be a resource for the symbolic construction of a variety of dual structures. Its use, however, is selective and should not be confused with these forms of dualism. Finally Chapter 8, written and eventually published in 2010, deals with the use of ‘formulaic’ expressions in Rotenese ritual language. In it, I examine a few of the formulae used by five different poets in composing some 18 or so lines all excerpted from single chant Suti Solo do Bina Bane, which I first began recording in 1965. Although this paper is merely illustrative of the use of such formulae, it could only have been written after extended examination of numerous compositions over many decades.

The second section of this volume goes on to examine the traditional Rotenese oral canon, knowledge of which can only be conveyed in ritual language. This is a rich canon that has taken many years to compile and fathom and has involved numerous return visits to Rote since 1965. Although I have recorded ritual language compositions from all of the main dialect areas of Rote, over the years, I have concentrated on an in-depth recording of the ritual language of the speech community of Termanu. Most of the traditional oral canon cited in this volume comes from this single speech community and reflects my growing personal involvement in the social life of that domain. Based in Termanu at the centre of the island, however, I have been able to look both east and west in my research.

The five chapters that comprise this second section provide a glimpse—but only a glimpse—of what was once regarded as most venerated of Rotenese ancestral knowledge. Much of this knowledge has to do with mythic relations between the Sun and Moon and the Lords of the Sea. Chapter 9 was written as an attempt to sort out some the genealogical relationships among these heavenly ancestral figures as means of ordering the different parts of the Rotenese canon of origin chants. Chapter 10 examines the oldest recorded specimen of ritual language, a text originally recorded by the Dutch linguist, J.C.G. Jonker, probably in the late nineteenth century and published without translation in 1911. In translating and commenting on this beautiful composition, I have tried to identify some of recurring themes that dominate Rotenese conceptions of human life. Chapter 11 examines a particular origin chant that recounts the coming of the seeds of rice and millet to the island and their dissemination throughout the fields of the domains. Just as persons, so, too, places on Rote are given dual names in ritual recitations. Each part of the island of Rote—its many domains, individual fields and gardens, and its innumerable significant sites—possesses at least one and sometimes a succession of dyadic ritual names. In Chapter 11, I introduce the ideas of ‘topogeny’ as an ordered recitation of place names that serves, as does the ordered succession of names in a genealogy, to provide a narrative structure. Not only is the use of topogeny a common feature in Rotenese ritual language compositions, it is also a common compositional feature in other traditions of parallelism. Chapter 12 offers a brief examination of another chant that recounts the origin of a specific variety of millet.

Mortuary chants form a large component of the traditional Rotenese canon. Chapter 13 provides an example of a mortuary chant of particular interest in that the deceased is given voice at his own funeral to admonish the living before departing for the land of the dead. Chapter 14 is an account of the mortuary rituals that I sponsored for the ‘Head of the Earth’—Stefanus Adulanu, known as ‘Old Meno’—who was my teacher and collaborator in my initial efforts to learn ritual language in 1965–66. The Chapter examines the importance of the final mortuary memorials of ‘rock and tree’ erected to notable individuals and discusses the significance that Rotenese accord to the ‘aroma’ of the name that lives on in memory long after death.

The third section of this volume looks at how, under the influence of Christianity, another sacred canon has been created and is now, appropriately, expressed in ritual language. A crucial factor in Rote’s history was the adoption of Christianity by various domain rulers in the early eighteenth century. The adoption of Christianity was closely linked to the use of Malay and particularly to the use of the Malay Bible. This led to the early establishment of local domain schools, creating a Malay-speaking, educated elite on the island and more widely in the Timor area. Malay, in various forms, became an accepted speech register on the island—the only speech register that was considered, for at least 150 years, as the proper vehicle for writing and the only appropriate vehicle for preaching the Scriptures.

This situation began to change towards the end of the nineteenth century. Increasingly, in the twentieth century, parallel compositions in ritual language became the vehicle for Biblical knowledge among a new cohort of local Rotenese preachers. Three chapters in section three give some idea of these changes. Chapter 15 discusses the historical creation of this new Christian oral canon and the appropriation of Biblical knowledge for its creation.

The degree to which individual recitations conform to Biblical foundations varies. At present, there are many compositions in ritual language that merge traditional and Biblical forms of knowledge. Chapter 16 provides an example of one such recitation that in its retelling of the story of Adam and Eve purports to recount the origin of death. Finally Chapter 17 gives an example of the use of parallelism in a Rotenese sermon. As knowledge of the traditional oral canon declines, the use of ritual language is becoming ever more important for the performance of Christian ceremonies.

In 2006, I embarked on a renewed effort to understand ‘Rotenese ritual language’ more systematically across the entire island. Instead of attempting to ferret out particular Rotenese ritual-language specialists in their local settings, I began by inviting poets from different dialect areas to Bali for week-long recording sessions. I relied on a few leading poets to select other poets whom they knew either personally or by reputation. Bali was strategically chosen as a gathering point specifically to remove, or at least lessen, the various cultural requirements that restrict local recitation. For the poets, the plane trip (‘above the clouds’) to the ‘Land of Bali and the Water of the Gods’ (Bali Dae//Dewata Oe) provided a separation from Rote that was liberating. Together in a group, with poets from other parts of Rote, most of the poets who came to Bali were intent on demonstrating their domain’s special linguistic heritage.

To date, I have held eight recording sessions on Bali between 2006 and 2013 involving 27 different Rotenese poets from 10 separate domains on the island. The initial objective of the project, which continues, was to gather as wide and as rich a corpus of ritual compositions as possible from the full spectrum of Rotenese cultural traditions among the different domains and from as many distinct dialects as was possible. The past several years have been perhaps the most fruitful period of research on Rotenese parallelism.

The last chapter in this volume discusses this research in some detail. It serves therefore not so much as a conclusion to this volume as a kind of on-going discussion of current and future research directions for the comparative study of semantic parallelism.

Theme and idiom in the two oral canons

The distinction that I make between a traditional oral canon and a Christian oral canon points to an emerging difference in the nature of ritual language compositions and, more significantly, a difference in a general regard for foundational knowledge in Rotenese culture. These changes reflect the Rotenese conversion to Christianity but more importantly the assimilation and appropriation of Christian ideas within Rotenese culture. They also reflect a change in language usage: a switch from an exclusive reliance on Malay as the vehicle for Christian preaching to a greater use of Rotenese in combination with Malay. I examine these historical changes in Chapter 16.

Although these distinct canons are intended to convey different sources of knowledge, the idioms and cultural concerns of the two canons remain much the same. Thus, for example, the two canons are deeply concerned with origins: the beginnings of things and their consequences. In the traditional canon, the most important recitations constitute a special set of origin narratives, whereas in the Christian canon, the most frequent recitations involve the retelling of Genesis. Both Chapter 15 and Chapter 16 give examples of these retellings.

Similarly, in both canons there is a great emphasis on speaking—a continued internal dialogue among the actors of the narratives. In the traditional canon, the Sun and Moon direct their children, their sons argue with the Lords of Ocean and Sea, the birds of the air and creatures of the depths oppose, object and dispute with their interlocutors, spiders and stick-insects announce their creative actions, and in the funeral chants, the dead are given leave to instruct and admonish the living. In the Christian canon, God speaks His creation into being; He speaks to Adam while Eve speaks with the serpent. Speaking, for the Rotenese, is the indispensable feature of sociability. Both canons reflect this cultural obsession.

Throughout both canons, there is also a continuing reliance on a botanic idiom to describe human development and personal interrelations. Botanic icons—a host of different plants and trees, each with its particular characteristics—are used as images for growth and differentiation, for the spread of knowledge and for the paths of life. The world may be metaphorically represented as a great banyan tree with branches leading in different directions; trees whose branches scrape together provide a prime image of social interaction; in rituals, boys are likened to banana and sugar cane; women to coconut and areca nut. Similarly, in the Christian canon, the establishment of Christianity is seen as the planting of a new tree, and its spread, like the spread of rice and millet recounted in the traditional canon, involves its dissemination and successful cultivation in new fields.

Another feature of both canons is the narrative recurrence of the journey. Chant characters move through an identifiable landscape and their journeys are significant. These journeys are undertaken in the search for companionship or for a spouse, for wisdom and special knowledge or simply for a resting place in Heaven or in the realm of the dead. Journeys proceed through a succession of named places whose exegesis is often vital to an understanding of the quest. Many Rotenese ritual-language narratives follow a structure of ordered locations or more explicitly recount an ordered succession of placenames—what I call a ‘topogeny’ that has similar significance to the personal succession of names in a genealogy. For the Rotenese, life is a journey and their rituals proclaim this view.

A more specific image but one that is central to this world view is the idea of the ‘orphan and widow’, which I discuss in Chapter 10. Virtually all the mortuary-ritual recitations of the traditional canon involve this image in concrete terms, but underlying the specifics of mourning and separation is the more general understanding that death, as the fate of all humanity, is the one journey that is common to everyone and thus the denominator that reduces everyone to the same status. This view—that everyone is an orphan and a widow—is enunciated in both the traditional canon and the Christian canon.

Personal engagement

There is also a deeply personal dimension to this research. The essays in this volume represent research that was begun in the 1960s and has continued to the present. This research has involved a continuing personal engagement with master poets whose oral compositions I have been able to record and discuss in my attempts to translate and comprehend them. This personal engagement has figured prominently in my work and is clearly articulated in the chapters that make up this volume. Most of these poets I consider as friends. Some have been my closest friends on Rote.

Chapter 4 for example, contains a composition entitled Dela Koli ma Seko Bunak by the poet Stefanus Adulanu, who was known simply as Old Meno (Meno Tua) at the time of my first fieldwork. In 1965–66, Old Meno took me under his wing and became my principal teacher of ritual language. Elsewhere I have written at length about my involvement with Old Meno. Here it is perhaps enough to quote some of my earlier comments on the way he coaxed and guided me to an understanding of ritual language:

[Old Meno] became for me my most profound mentor. He had a way of teaching which I found at first frustrating but I soon came to realize that it was—and is—the most essential way to teach. Meno would ‘guide’ me to an understanding but never make explicit what he was guiding me toward. Thus it was always up to me to make connections and come to my own understanding. Much of our time together involved circling an understanding and only when he felt that I had achieved some degree of comprehension could we move on further…Learning from Meno was one of the most valuable experiences of my life. (Fox 1992:109)

Old Meno died before my first return visit to Rote so I was only able to record him in 1965–66, yet in my corpus, Old Meno’s compositions are among the finest and most significant pieces I have recorded. Many of them are origin narratives that were, at the time I recorded him, considered ritually circumscribed and inherently powerful. On one occasion, Old Meno told me explicitly that he had left out a key passage from a narrative but he also made it clear to me what that passage revealed. The composition Dela Koli ma Seko Bunak is in fact linked to the restricted corpus of origin narratives. When Old Meno recited this chant, he left out its origin dimensions and instead presented it as a mortuary chant. Years later, another chanter, Seu Ba’i, who had actually heard Old Meno recite this chant for me, revealed that it was in fact the chant that recounted the origin of two prominent rocky outcrops, Sua Lai ma Batu Hun, that are considered defining natural icons of the domain of Termanu. Seu Ba’i gave me his version and I have subsequently recorded yet another version of this chant.

In addition to Seu Ba’i, Peu Malesi was another Termanu poet from whom I recorded a great deal of material. In 1965, he was the first poet who dared to provide me with my first recording of the origin of fire and cooking, excerpts of which are included in Chapter 6. On my second long stay on Rote, in 1972–73, he visited the house where I stayed at Ufa Len so frequently that he became my regular companion and drinking partner.

Peu Malesi had great versatility in his compositions. Chapter 11 contains his narrative of the origin of rice and millet together with his extended topogeny of the spread of these seeds throughout Rote; Chapter 15 contains his account of the origin of death, a version of Adam and Eve in a Rotenese setting. Importantly as well, towards the end of my second visit, Peu Malesi took a prominent role in the performance of the mortuary ceremonies that I sponsored in honour of Old Meno. In 1977 and again in 1978, when I visited Rote with the filmmaker Tim Asch, we were able to film him chanting. He appears in both films that we made: The Water of Words (1983) and Spear and Sword (1988). Although Peu Malesi had great abilities, he was unable to replace Old Meno as a teacher. Peu Malesi could recite at great length but could not provide any kind of reflective exegesis on his own compositions. Often I would work with him to transcribe one of his chants. Invariably he would refer to the voice on the tape recorder as an externalised third person. His response to most of my questions was to recite another (often similar) passage.

During my first fieldwork on Rote, I spent time in Oe Handi in the domain of Thie, living briefly in the house of the oldest and most renowned poet in that domain, N. D. Pah, who was known as Guru Pah. His first question to me when I arrived at his house was why it had taken so long for me to come to see him. The origin chants that he recited for me were as revelatory as those that I had recorded from Old Meno and Peu Malesi. These chants and subsequent chants that I recorded in Thie also provided me with another perspective on the ‘traditional oral canon’ as I have described it: basically, that this canon may once have constituted a more coherent repertoire of recitations recounting relations between the Sun and the Moon and the shark and crocodile as Lords of the Sea. What I was encountering in my fieldwork were separate fragments of a large whole. This is an idea that I develop in Chapter 9, ‘Genealogies of the Sun and Moon’, and it underlies my continuing efforts to gather chants across dialect areas of the island.

Yet another poet of considerable ability whom I was able to record in 1965–66 was Stefanus Amalo, who was almost the same age as Old Meno. I was able to visit him only once in Pua Mata, where he lived, but he came to visit me in Ufa Len. One evening, close to midnight, he suddenly arrived and told me he wished me to record his recitation. This recitation, Pau Balo ma Bola Lungi, came to roughly 390 lines. I have since recorded other versions of this chant but none that compares in length or beauty with this first version.

The person who became my closest friend on Rote and eventually my most reliable ritual-language commentator was Esau Pono, whom I hardly knew in 1965–66. At that time, he was a young man, only a few years older than me, and was involved in a difficult dispute over the payment of bride-wealth to marry a woman from a clan that was based in Sosadale near Ufa Len. Six years later, by the time of my second visit, he had begun to gain a reputation as a local preacher and I turned to him for his help to provide the Christian component to the traditional ceremonies that I held in Old Meno’s honour. On subsequent visits, we became good friends. Thus, as part of our film work in Termanu, Tim Asch filmed ‘Pak Pono’, as he was known, conducting a Sunday service in the local church near Ufa Len. Afterwards, we invited him to Canberra to help with transcriptions and translations.

In the years that have followed, I have kept in close contact with Pak Pono. During this time, he has become Termanu’s most respected ritual chanter. In terms of knowledge, fluency and versatility, he is without a rival within the domain and is well known among the chanters of the island. Therefore when I embarked on my project to record poets from all the dialect areas of Rote, I recruited Pak Pono to join me in my effort and he has continued to participate in each subsequent recording session, helping also to select and invite other notable chanters. In the process, I have been able to record from him two separate recitations of the chant Suti Solo do Bina Bane, but more importantly still, I have relied upon his deep knowledge of his language and traditions. Like Old Meno, he is able to reflect and comment on his own compositions and those of others.

Figure 1.1: Esau Pono, poet and preacher, my oldest friend and collaborator on Rote

As part of new research on the dialects of Rote, I have come to know at least another dozen master poets. Among these poets, there is one who stands out. This is Ande Ruy from the domain of Ringgou. At present, ‘Pak Ande’ is the most able poet and performer on Rote and is generally recognised as such. Pak Ande is a born showman: he sings, he drums, he plays the Rotenese sasandu and he is an exceptionally fluent chanter in a variety of modes. As a consequence, he is frequently called upon for official cultural performances.

Pak Ande and Pak Pono, to whom he defers, have participated in all the recording sessions I have held. Pak Ande has provided me with a torrent of material, which I have struggled to understand because of the density of its metaphors and allusions and because it is in a dialect that I find most difficult to comprehend. Even Pak Pono has admitted to me that he, too, has difficulty at times in understanding Pak Ande. The second of the Genesis recitations in Chapter 16 is a good example of a Pak Ande composition. In many ways, Pak Ande is like Peu Malesi. The exegesis that he offers of his recitations often involves more recitation.

All Rotenese chants are bound in ritual. They are dense revelations that require a degree of exegesis (and possible paraphrase) to make them intelligible. When I began recording these chants, particularly the origin narratives, they were considered as restricted knowledge. For that reason, I have hesitated to publish in full my main corpus of origin narratives. Chapter 6 in this volume provides a selection of passages from the narrative of the origins of fire and of cooking with appropriate commentary, but not the whole of the narrative. Similarly, I have published passages from the narrative of the origin of weaving in the paper ‘Figure shark and pattern crocodile’ (1980), and various passages from the narrative of the house in a paper, ‘Memories of ridgepoles and crossbeams’ (Fox 1993).

Over the past 45 years or more, the restrictions on Rotenese chant revelations that once applied have gradually lifted. I have discussed this issue with many poets, including both Pak Pono and Pak Ande, and all of these poets have assured me that it would now be appropriate to publish my corpus of origin narratives, particularly the long versions of the chants that I recorded in 1965–66. This represents yet another project that I have before me. As I see it, there remains much to be done before I can feel satisfied in having presented the full richness of Rotenese ritual language.

A note on the transcription of Rotenese

As I explain in this volume, Rotenese rulers established local Malay-based schools in their different domains beginning early in the second quarter of the seventeenth century. Until nearly the end of the nineteenth century, all writing was confined to Malay; however, this began to change in the twentieth century, as the Rotenese—particularly schoolteachers (known as ‘Malay Masters’)—began writing their own language. For this purpose, they relied on the same Latin script used for Malay. Various attempts were made to modify this script, but in the end, the same script used for Malay was used for Rotenese. This was not a perfect solution because Rotenese dialects have sets of long vowels as well as short vowels (e, o, u, i), but it was adequate for communication and did not jeopardise understanding. Except for a few words, the contrast between long and short vowels does not produce phonemic distinctions. Moreover, there are regular patterns to the use of these long and short vowels. Thus, for example, the second (and third) vowels in a sequence of similar vowels (as in partial reduplication) are lengthened: dedéak, kokólak, lulúnú, titítík. The transcription of Rotenese used in this volume follows common Rotenese usage.