https://orcid.org/0000-0001-9295-5741
![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
Abstract
Sixteen hillforts constructed and occupied perhaps several centuries before contact in the early nineteenth century are described from Bua district in northern Fiji. These hillforts represent inland settlements in fortifiable locations on high volcanic islands, plausibly established in response to the outbreak of sustained conflict. The chronology and functions of these hillforts were investigated through mapping, excavation, and collection of oral traditions. Four groups of hillforts are recognized. The Seseleka and Yadua (Island) groups represent single polities, comprising a mountaintop site surrounded with fortified narrow steep-sided basalt ridges with tributary sites at lower levels functioning as lookouts and/or food-processing sites. The Northwest Bua group comprises mountaintop sites occupied only when aggressors threatened, and are otherwise characterized by occupation around their bases. The Inland Bua group ranges along one of the highest ridges in the area and were all reportedly in conflict with each other. The Buan hillforts represent those found elsewhere in Fiji and on other high Pacific Islands. Insights from the study of these hillforts illuminate this period of Fiji history and help interrogate broader questions about drivers and proximate causes of the conflict that may have seen hillforts established near simultaneously on high Pacific Islands during the last millennium.
Acknowledgements
PN is grateful to the University of the Sunshine Coast for funding his initial collaborative research in Bua with the Fiji Museum and the University of the South Pacific. Later support through the Australian Government’s New Colombo Plan helped fund the three phases of research in Bua. In-kind support was provided by the Fiji Museum, the Fiji Government’s Ministry of iTaukei Affairs, the Commissioner Northern’s Office, and the Bua Provincial Council. MM received funding for three radiocarbon dates from the University of the South Pacific. Most of all, the researchers are grateful to the people of Bua who were very supportive of this research; oru sa vakavinavinaka vakalevu vei kemiau a turaga kaya marama ni vanua vakaturaga ko Cakaunitabua ina omiau veivakaitaukeitaki kaya veitokoni ina sasaga ni vakadidike qoi. We especially acknowledge the late Matanitikina of Bua, Nevote Roko, who selflessly facilitated our introductions to the people of Bua, and Sepeti Matararaba (Fiji Museum) who was involved in early fieldwork in Bua. Posters summarizing the results of this research have been prepared in English, Fijian (iTaukei Bauan and Buan), and Fiji Hindi and have been distributed among the schools and communities in the study area; pdfs are available free from PN.
Notes
1 Today koro is used for all iTaukei settlements in Fiji, coastal or otherwise. The earliest use of koro in this sense was in 1840 (Wilkes Citation1845), suggesting its original meaning was then already in the process of transformation.
2 Bua Bay (in Figure 1) was named Sandalwood Bay on the earliest maps of the area
3 The highest parts of the hillforts of Devoka and Uluinasiva proved impossible to adequately access within the time frame of the project.
4 The meaning of Seseleka is uncertain but the most likely option is that it means “nearby [place of] refuge” (Seesee-leka).
5 Note that Wilkes’s use of Dimba-dimba was probably because local people referred to Naicobocobo Point as icibaciba (pronounced ee-thimba-thimba) or the place of the dead, an allusion to the role that Naicobocobo had as the stepping-off point of dead souls.
6 This was noted ethnographically elsewhere in Fiji (Clunie Citation1977).
7 This story is widespread today as well as in 1919 when it was reported that local people believed the pool “to be connected underground with the sea; the tradition being that the water was at one time salt and rose and fell in sympathy with the tides” (Parham Citation1922:8).
8 Almost certainly, the overlord was the Tui Bua, based as today in the chiefly village of Bua (Lomanikoro).
9 The literal meaning is “big sea” or “big coast”, perhaps in the sense of this hilltop commanding a broad view of the coastline, which explains why it was used as a hillfort, especially one tributary to higher Seseleka where mist/cloud may often have contributed to a less-complete view of the coast.
10 This a descriptive name for the breccias that cap this peak and account for its prominence in the landscape. The breccia gives rock faces a rough (not smooth) appearance and can be abrasive, hence (a place of the) rock (vatu) with scabies (karokaro).
11 This name probably means “a meeting house with eight [supporting] posts”; na is the (meeting house is implicit), bou is post, walu is eight.
12 Since the literal name for “earthen wall” in Fijian (baiqele) is not used at Nabouwalu, this feature is more likely the remains of the foundations of a fence of wooden stakes, bamboo, or reeds. Such palisades are noted in a few observations of nineteenth-century Fijian hillforts (e.g., Lockerby Citation1925) and were defining features of pā, their New Zealand counterparts (Schmidt 1996).
13 Osoabukete (or Osonabukete) is a quite common place name in Fiji for places that can be accessed only through a narrow (or tight) space. Literally meaning “too narrow for a pregnant person”, at this location there is said to be a place between two rock faces through which one has to squeeze, but it was not located during fieldwork. Osoabukete Cave on Lakeba Island (eastern Fiji) is described elsewhere (Nunn et al. 1991).
14 Although vatu means rock in Fijian, the truncated nature of this name makes it difficult to be sure of its meaning. A plausible meaning is “rock shaped like a tuber”.
15 So named by Captain William L. Hudson, commanding the Peacock, one of the ships of the U.S. Exploring Expedition, on June 11, 1840, on account of the resemblance of the rock outcrop from the sea to a monkey (Wilkes Citation1845). Although there are no monkeys in Fiji, the name is now common.
16 This may have derived from the scorched or parched nature of its treeless summit.
17 Savage’s group subsequently found themselves on Dillon’s Rock (see Figure 9, right), surrounded by hostile Fijians, who eventually caught and killed him (Campbell 1980).
18 Possibly the etymology of Nasiva derives from “na” (the) “siva”, the Fijian name given to several species of oyster-like bivalve (including species of Isognomon, Pinctada, and Pteria). Perhaps the leader of the people who once resided on top of this mountain wore, as high-status people in pre-contact Fiji were known to wear, a necklace with one or more large siva threaded on it.
19 The meaning of the name Talai is uncertain but the possibility that, as in certain Polynesian languages, it means “adze” and is an allusion to a nearby quarry (not located) from which adzes were once made appears quite strong.
20 The name means “Kota Mountain” and it is uncertain whether kota is a name or refers to a tree (perhaps a species of Dysoxylum) that once grew there.
21 Based on 65 radiocarbon dates (Nunn and Petchey 2013), the average 1-m-thick shell midden at Bourewa (Fiji) accumulated within 162 years, possibly 299 years, suggesting that in a comparable situation (Koromakawa on Yadua Island), a 1.7-m-thick midden may have accumulated in 275 years, perhaps 508 years.
