Previous Next

Master Poets, Ritual Masters

12

Historical Diversity and Dialect Differences on Rote

A Brief History of the Rotenese Domains (Nusak)

Rote is a relatively small island located off the south-western tip of West Timor. The island measures roughly 83 km in length from south-west to north-east, while, at its widest point, it is 25 km across. It covers an area of approximately 1,670 sq km, which includes a number of tiny islands scattered along its coast. Rote is a low-lying island with stretches of savannah and occasional limestone hills on its southern flank.

Despite its diminutive size, Rote is remarkably diverse—linguistically, socially and in its political history. In the mid-eighteenth century, after a number of devastating reprisals for opposing the Dutch presence, local rulers on Rote sought to establish close relations with the Dutch East India Company. In 1662, the first group of four local rulers gained Dutch recognition of their domains. These were the rulers from Termanu, Dengka, Bilba and Korbaffo. Within a few years, in 1690, another eight local rulers officially pledged their loyalty to the Company. This second group came from Landu, Ringgou, Oepao, Bokai, Loleh, Lelain, Thie (Ti) and Oenale.

Figure 13: Map of the Domains (Nusak) of Rote

Source: © The Australian National University CartoGIS

Thus, even before the beginning of the eighteenth century, Rote was divided among a dozen small polities, each with its own ruler and ruling dynasty. But this did not put an end to the creation of further polities on the island. Indeed the Dutch archives of the period mention quite a number of local areas clamouring for recognition and their own political autonomy. The first of these was the domain of Ba’a, which was split from Lelain by Dutch agreement in 1700. Also mentioned at the time was the domain of Diu, which only succeeded in obtaining recognition in 1756. Following Diu, the domains of Lelenuk, Keka and Talae were separated from Termanu in 1772.

Whereas at the beginning of the eighteenth century, Rote was divided among 13 polities, by the nineteenth century, the number of these official domains had increased to 17, with yet another domain, Delha, to be split from the larger domain of Oenale and to be given official recognition sometime in the early nineteenth century. At this point, the Dutch policy of fostering ever more divisions on the island came to an end. Instead, under colonial government policy, dissident local populations were transported to Timor to form a buffer around the Dutch settlement at Kupang.

In the twentieth century, the Dutch Government attempted, for purposes of governance, to amalgamate various domains. These efforts did not lessen the social reality of the separate domains nor did the Indonesian Government when initially it made the whole of Rote a single political entity (kecamatan). With the policy decentralisation after 2002, Rote has become a higher-order political entity (kabupaten) and has been divided into a number of kecamatan. Among these new kecamatan, the large domains of Landu, Termanu, Dengka and Thie have been accorded autonomy, while other domains have been grouped with neighbouring domains. The domain structure established in the seventeenth century continues to this day on Rote.

On Rote, each of the domains originally recognised by the Dutch is referred to as a nusak. Each nusak was and remains the focus of long-standing social identity. Although Dutch recognition may go back more than 350 years, the oral narratives of these domains claim a distinct ancestral past that extends well before the arrival of the Dutch. These oral narratives recount the coming of the earliest ancestors but also successive developments in the formation of the domain. For most domains, this has been a development of relations among the constituent clans of the domain.

For one domain, Landu, once among the largest of Rote’s domains, this development was aborted. In 1756, in retaliation for opposition to the East India Company, Landu’s population was devastated. Some of its population was able to flee while others were either slaughtered or sold into slavery. For decades thereafter, Landu became a no man’s land. When in the nineteenth century Landu was gradually resettled, it never regained its previous integrity. By contrast, despite dismemberment in 1772 of the domains of Talae and Keka on its southern coast, Termanu retained a privileged position on the island, occupying a considerable area of central Rote with a port on its north coast that was used by the Dutch as the primary entry to the island. From the beginning of relations with the Company, the Dutch tended to deal with Termanu among the domains of Rote.

Each domain on Rote was a diarchy with rule divided between a superordinate ‘male’ lord (manek) and a subordinate ‘female’ lord (fettor). Each domain was defined by its local court, situated in the chief settlement of the male lord. This settlement as the centre of the domain with the court as its focus was also referred to as nusak. The court was presided over by the male ruler, who was supported by the heads of the clans of the domain. Disputes were regularly heard at court and decided on by the ruler in deliberation with its clans (see Fox 2007b). This dispensation of traditional justice based on local customary practice provided a focus of intense interest and a prime forum for argument, debate and litigation—the disputation and taking sides that, for Rotenese, are among the great joys of life.

All domains are made up of clan groups that are distinctive to that nusak. There is no island-wide network of clans to unite the island. Clans are arrayed as either noble or commoner, with one or more clans recognised as the original ritual custodians of the earth. The ordered arrangement among clans differs from domain to domain. Termanu, for example, has a subtle hierarchy of lineages particularly within its large royal clan. In turn, this royal clan is linked to the other clans of the domain according to their derivation or initial affiliation and incorporation within the domain. By contrast, all the clans of Thie are divided into one or another moiety: Sabarai and Taratu. Sabarai was presided over by the male lord and Taratu by the female lord (see Fox 1979a, 1980). Clan groups intermarry with one another so that the overwhelming majority of all marriages are contracted within one’s domain.

All nusak were also, once, distinct ritual communities whose integrity was affirmed in a series of ‘origin’ ceremonies performed by the constituent clans of the domain. Ceremonies of the life cycle were also performed in a distinctive manner. These ceremonies—of birth, marriage and death, along with the rituals associated with house-building, tie-dyeing, weaving and planting and harvesting—were carried out by reference to chants that invoked the ‘origin’ of these activities.

Changes to the ceremonial life of the nusak began in 1729 with the conversion to Christianity of the first ruler of a domain, the domain of Thie. This was followed by the conversion of other rulers from smaller domains in south-central Rote, who together claimed Dutch protection as sovereign Christian states. Coupled with conversion came a request for Malay schoolteachers to establish nusak-schools. By the middle of the eighteenth century, the prestige of having a school and the rivalry among the nusak led to the establishment of numerous schools, but the costs demanded by the Company eventually favoured the larger domains, most of which did not have Christian rulers (see Fox 1977: 106–12).

In time, a tradition of schooling in Malay and of adherence to Christianity became part of Rotenese life. Schooling and Christianity went hand-in-hand but both progressed slowly. Towards the end of the nineteenth century, Rote had 34 local schools with more than 3,000 students, but after more than 170 years of Christianity, only one-fifth of the population of the island was estimated to be Christian. In the words of a Dutch missionary, the Rotenese wore Christianity as ‘Sunday apparel’ over their heathen interiors (see Fox 2014: 321 ff.).

Only in the twentieth century were there dramatic changes to the ceremonial lives of the different nusak on Rote. These changes came with the use of Rotenese and Rotenese ritual language, instead of Malay, for the preaching of Christianity. Gradually, the nusak abandoned their origin (hus//limba) ceremonies. Thie, the first domain to become Christian, was one of the last to cease to perform its origin ceremonies. Today one village in Dengka still performs an origin ceremony but this ceremony is limited to a small community of that village.

Similarly but more slowly, the ceremonies of the life cycle began to change and were either combined with or gradually replaced with Christian ceremonies. In the process, ritual language developed a second lease on life and an extensive new vocabulary as it was used increasingly in the services of the Protestant Church, particularly in sermons and especially in rendering the Bible into elevated Rotenese. As a result, the knowledge of origins and the memory of many traditional rituals exist alongside the creative use of ritual language blending in a Christian vein.

The Languages of the Different Domains (Nusak)

Throughout Rote, it is categorically asserted that each nusak has its own ‘language’. Superficially this is undeniable. The local geography of each domain is unique: place names and clan names are distinctive as are other identifying features of the domain. Behind the assertion about the different, distinct languages of the domains lies the recognition of considerable dialect diversity across the island.

Rote consists of a dialect chain of related languages. Speakers in neighbouring domains are generally able to understand one another, but for speakers in domains separated from one another intelligibility is reduced. Domains at a distance from one another find mutual intelligibility difficult or impossible. Based on these criteria, Rotenese consists of more than one language.

Various attempts have been made to distinguish the dialects of Rote. In the nineteenth century, a Rotenese schoolteacher, D. P. Manafe, proposed a grouping of nine dialects based mainly on phonological (Indonesian: lagu) criteria. His grouping, based on a native acquaintance with the languages of the island, is pertinent.

His group of nine dialects is the following:

1)

Eastern Dialect:

Landu, Ringgou and Oepao

2)

East-Central Dialect:

Bilba, Diu and Lelenuk

3)

Korbaffo Dialect:

Korbaffo

4)

Central Dialect:

Termanu, Keka and Talae

5)

Bokai Dialect:

Bokai

6)

South-Central Dialect:

Ba’a and Loleh

7)

North-East Dialect:

Dengka and Lelain

8)

South-West Dialect:

Thie [Ti]

9)

Western Dialect:

Oenale and Dehla

Despite decades of research on the Rotenese language, the Dutch linguist J. C. G. Jonker never ventured a definitive grouping for the dialects of Rote. When he published his Rottineesch-Hollandsch Woordenboek (1908) based on the Termanu dialect, he was acutely aware of dialect variation. As a consequence, for many Termanu lexical items, he included variant forms listed according to specific domains. At the end of his dictionary, he also included a long list (103 pages) of ‘alternative forms and words from other dialects’. Subsequently, he published an extended article on the dialects of Rote, ‘Bijdragen tot de kennis der Rottineesche tongvallen’ (Jonker 1913). Most of the article consists of similar texts in different dialects, but as an introduction, Jonker offered comments on the dialect situation on Rote. His comments, which begin with a consideration of Manafe’s earlier assessment of dialects, are instructive but inconclusive. For the most part, Jonker simply summarises the variety of phonological differences among dialects. He disagrees with some of Manafe’s distinctions—on the one hand, pointing to differences among dialects that Manafe grouped together, and on the other hand, noting the continuous variations among dialects that Manafe separated. A major point that he does emphasise is that putting aside phonological criteria and looking instead at grammatical criteria, there is a significant difference between the languages of Dengka and Oenale in the west and the languages of eastern Rote, with Thie situated in a kind of intermediary position. Many of the distinctive features of Dengka and Oenale are similar to those of Timorese.

When I embarked on my attempt to record and translate ritual language compositions across the dialects of Rote, I had to my advantage the previous work by Manafe and Jonker, but not until I was already launched into my recording sessions did I come to appreciate the complexity of variation among Rotenese dialects. As a basic aid to my own understanding, I drew up a list of the principal sound variations in the main dialects that I came to focus on. Although this list is by no means complete, it provides a starting point for comprehending the sound variations one encounters in the ritual languages of Rote.

Table 5: Principal Sound Variations in Rotenese Dialects

Termanu values:

Ringgou

Bilba

Termanu

Thie

Dengka

Oenale

Initial

0

k

k

k

0

0

[k]

ona

kona

kona

kona

ona

ona

Medial

0

k

k

k

k

k

[k]

se’o

seko

seko

seko

seko

seko

Initial

d

d

d

d

l

r

[d]

dale

dale

dale

dale

lala

rala

Initial

k

ng

ngg

ngg

ngg

ngg

[ngg]

kia

ngia

nggia

nggia

nggia

nggia

Medial

k

ng

ng

ngg

ngg

ngg

[ng]

boki

bongi

bongi

bonggi

bonggi

bonggi

Initial

r

l

nd

nd

nd

nd

[nd]

rai

lai

ndai

ndai

ndai

ndai

Medial

n

n

n

nd

nd

nd

[n]

tane

tane

tane

tande

tande

tande

Initial

b

b

b

b

f

f

[b]

bafi

bafi

bafi

bafi

fafi

fafi

Initial

p

p

p

mb

mb

mb

[p]

peda

peda

peda

mbeda

mbeda

mbeda

Medial

p

p

p

mb

mb

mb

[p]

hapu

hapu

hapu

hambu

hambu

hambu

Initial

l

l

l

l

r

l

[l]

ledo

ledo

ledo

ledo

lelo

ledo

r

l

l

r

l

r

Medial

r

l

l

r

r

r

[l]

hara

hala

hala

hara

hara

hara

Note: In relation to Termanu dialect, 0 = ‘absence of’.

As a further means of distinguishing these dialects, I drew up another chart of the pronominal system of the different dialects. This chart gives some indication of Jonker’s observation about differences between eastern and western Rote.

Table 6: Pronominal Systems of the Different Dialect Areas on Rote

I

II

III

IV

V

VI

Ringgou

Bilba

Termanu

Thie

Dengka

Oenale

1st p. sg.

au

au

au

au

au

au

2nd p. sg.

o

ko

o

o

ho

ho

3rd p. sg.

ria

ndia

ndia/ana

ana

eni

eni

1st p. pl. (excl.)

kita

kita

ita

ita

hita

hita

1st p. pl. (incl.)

ami

ami

ami

ai

hai

hai

2nd p. p.l

emi

kemi

emi

ei

hei

hei

3rd p. pl.

ara

ala

ala

ara

ala

ara

Finally, on the basis of my recordings and my understanding of the dialects I recorded, I ventured to draw a map to provide a geographical dimension to distinguish among the dialects of Rote. This map in many respects follows Manafe’s language groupings. To produce it, I have ‘lumped’ related dialects rather than ‘differentiating’ them in the attempt to reduce Manafe’s nine dialect groups to six, each of which is focused on a major domain. This reflects the fact that in my recording efforts, I was unable to find poets from many of the small domains of Rote and had therefore to rely on poets from larger domains.

This map distinguishes six ‘Dialect Areas’:

  1. An Eastern Dialect Area centred on Ringgou but including much of Landu and all of Oepao. A majority of the population to resettle Landu after it was depopulated in 1756 came from Ringgou and these Ringgou speakers are more densely settled on the south-eastern coast of the domain.
  2. An East-Central Dialect Area centred on the domain of Bilba but including the domains of Diu, Lelenuk and Korbaffo, with some influence on Landu because of migration to this domain.
  3. A Central Dialect Area centred on the domain of Termanu but including the domains of Keka and Talae with extended influence to Bokai and to Ba’a and Loleh.

    The Central and East-Central Dialect gives evidence of continuous variation among the different domains included within these areas. Hence the map shows no sharp line of demarcation across these two dialect areas. These two areas make up the most internally diverse yet related dialects of the island.

  4. A South-Western Dialect area centred on the domain of Thie.
  5. A North-Western Dialect area centred on the domain of Dengka.
  6. A Western Dialect area centred on the domain of Oenale including the domain of Delha, which was politically separated from Oenale in the nineteenth century.

Figure 14: A Tentative Map of the Dialects of Rote

Source: © The Australian National University CartoGIS

Dialect Diversity and the Semantics of Ritual Language

Rote’s dialect diversity is a prime resource for the elaboration of the island’s ritual language. Numerous word pairs in ritual language are composed of one term drawn from the local speech community and another term from some other dialect. These pairings are formal and are recognised as the established dyadic sets for that speech community. This is a process common to all the different speech communities of Rote, involving a combination of a local term and a dialect term, though not the same dialect term among the different domains.

What term belongs originally to the local speech community and what term derives from elsewhere can only be determined by reference to the usages within the local community. In a previous paper based on my earlier research centred on Termanu, I listed more than a dozen dyadic sets composed of a term from Termanu’s speech community combined with a dialect term from outside (Fox 1974: 80–83; 2014: 143–46). I also tried to specify the domain from which the ‘outside’ term originated and to categorise it as belonging either to eastern or to western Rote. In many cases, this designation was too domain-specific since subsequent research has indicated that ‘outside’ words generally have a wider ‘dialect area’ provenance. Nevertheless, a list of 10 of these dyadic sets is illustrative of the basic process of combining a common ordinary language term with one from outside that speech community.

Table 7: Dialect Terms in the Formation of Termanu’s Dyadic Sets

Termanu speech community

Outside dialect term

Gloss

lea(k)

lua(k)

‘cave, grotto’

li

nafa

‘wave, breaker’

henu

sofe

‘enough, sufficient’

lain

ata

‘heaven, sky, above’

longe

pela

‘to dance’

ka

kiki

‘to bite’

sele

tane

‘to plant’

lo

nggou

‘to call out loudly’

tenga

nggama

‘to take up, to grasp’

pu

oku

‘to shout, to scream’

A good example of dialect diversity is the dyadic set for ‘human being’, hataholi//daehena—an example that is iconic of the process of combining local and dialect terms. Hataholi is the term for ‘human being’ in Termanu and is regularly used as such in ordinary speech. Daehena, by contrast, is a dialect borrowing, which is considered to have come from eastern Rote. The linguistic reality is, however, more complex when seen from the perspective of the island as a whole. Each of the main dialect areas of Rote has a different variation of this dyadic set:

I

Ringgou

hataholi//lahenda

II

Bilba

hataholi//dahena

III

Termanu

hataholi//daehena

IV

Thie

hatahori//andiana

V

Dengka

hataholi//andiana

VI

Oenale

hatahori//andiana

Perhaps, interestingly, the distribution of this particular dyadic set is divided between eastern and western Rote with hataholi/hatahori as the common element in all dialect sets, though not the common local term for ‘human being’ in all dialects. The distribution of other dyadic sets across the dialects is not so neatly bifurcated.

In a closely related process, a speech community may adopt a pair of similar terms for a particular concept. Both may be terms that occur in that speech community, but given the dialect diversity on Rote, a combination of other pairs may be used in a different speech community but may, by analogy, be recognised widely across the island. Examples of these processes are numerous. Three particular instances—all of them of common dyadic sets—illustrate this process as seen from the perspective of the island as a whole.

In Termanu, the dyadic set for ‘name’ is nade//tamo. Nade is the general term for a person’s ‘name’ while tamo refers to a secret ancestral name, the name of a person’s ancestral guardian. Similarly, in Termanu the dyadic set for dry field or garden is tina//osi. It is composed of terms for different kinds of fields: osi is the term for a distant dry field while tina refers to a field in close proximity to the house. As a third example, the dyadic set tada//ba’e, which refers to the ritual distribution of goods, is composed of tada, which in Termanu dialect has the sense of ‘dividing’, and ba’e, which has the sense of ‘distributing’.

The distribution of the equivalent dyadic sets in the different dialect areas is as follows:

Table 8: An Illustration of the Concatenation of Dyadic Sets Across Dialect Areas

Dialect area

‘Name’

‘Field/garden’

‘To distribute’

I

Ringgou

nade//bo’o

tine//oka

pala//bati

II

Bilba

nade//bo’o

tine//oka

pala//bati

III

Termanu

nade//tamo

tina//osi

tada//ba’e

IV

Thie

nade//bo’o

tine//lane

ba’e//bati

V

Dengka

nade//tola

osi//mamen

pala//ndu

VI

Oenale

nade//nara

tine//osi

banggi//ba’e

The result of these processes, for the island as a whole, is a concatenation of pairs with varying components in different dialect areas. Importantly, however, whatever one’s position in the chain of Rotenese dialects, this concatenation of pairs represents ‘recognisable variation’ that serves two functions. It promotes intelligibility across divergent dialects and it enhances each dialect’s claim to be distinctive.

Dialect Versions of Suti Solo do Bina Bane

The second half of this volume is replete with a great variety of dialect concatenation. It includes nine separate versions of Suti Solo do Bina Bane composed by poets from all of the six dialect areas I have identified, from the eastern end of the island to its western end. The narrative of the chant varies far more than the versions of Suti Solo do Bina Bane in Termanu, as do the ritual contexts for these recitations. Perhaps most striking is the change of gender that the shells undergo from male in eastern and central Rote to female in western Rote. The shells also undergo a name change: in western Rote, Suti Solo do Bina Bane becomes known as Suti Saik do Bina Liuk.

As in the first half of the volume, in the second half the focus is on the parallelism of ritual language usage: the variety and continuity of formulaic expressions that continue to be maintained across the island of Rote. Despite phonological differences, there is clearly recognisable a core of shared dyadic sets that persists from one end of the island to the other. However, intermingled with this core of shared sets are distinct dialect variants that flavour each composition.

The nine versions of this chant come from the following master poets: Alex Mada of Landu and Anderias Ruy from Ringgou (Dialect Area I); Kornalius Medah from Bilba (Dialect Area II); Laazar Manoeain from Ba’a (Dialect Area III); N. D. Pah, Samuel Ndun and Jonas Mooy from Thie (Dialect Area IV); Simon Lesik and Frans Lau from Dengka (Dialect Area V); and Hendrik Foeh from Oenale (Dialect Area VI). Most of these compositions were recorded in gatherings on Bali as part of the Master Poets Project between 2006 and 2013. However, I gathered Laazar Manoeain’s composition during fieldwork on Rote in 1963 and the joint composition by the poets N. D. Pah and Samuel Ndun during fieldwork in 1973.

In the presentations of these various versions, I have tried to follow a roughly similar format. In presenting each composition, I provide background and exegesis to understand it. Following the translation of the chant, I initially try to identify the island-wide dyadic sets used in its composition—both those that are similar to the recognisable dyadic sets from Termanu and those sets that have undergone phonological change according to the dialect used in their composition. Thereafter I try to identify the dyadic sets and formulae that are distinctive to the dialect in question and compare these usages with usages in other ritual language compositions. At the beginning, comparisons are necessarily with ritual language usages in Termanu but once enough different dialect usages have been presented, comparison can be directed to a variety of dialect usages on Rote.


Previous Next