The most recent archaeological findings from Kosrae have prompted a re-assessment of settlement patterns, site distribution patterns, and interpretation of the cultural landscape. Our latest season of archaeological work was dedicated to a reconnaissance of the Tofol drainage, with the principal goal of locating Finol Tokosra, the King's Mound site. The site is well-known in the oral histories of the island, but the actual location of this legendary story of love and revenge had not been identified. During our reconnaissance, we located several sites throughout the drainage basin. At least two village sites hold promise as contenders for Finol Tokosra. Discussed here is our search for this site, the process of narrowing down the key elements in the oral histories as to what such a site would look like, what we actually found, and the integration of oral history and archaeology generally.
Archaeological interpretation is an exercise in tempered imagination. It relies on making reasoned connections between seemingly unrelated pieces of information and then generating multiple hypotheses rooted in the physical data-by that I mean the material residue that is the archaeological site and which stands as proxy for the events that ultimately formed the site. But these remains don't relay the whole story of the site; they simply establish one version. To flesh out the past, to anoint it with a humanity that is otherwise missing, we in Micronesia (well, some of us) make use of information derived from a variety of sources, including oral history.
For archaeology, the use of oral history has tended to be limited, focusing mainly on specific questions raised by the archaeological record. Such a concentrated use has its consequences (and I mean that in a good way), because it reminds us that cultural meaning and observation are firmly rooted in tradition . The caveat here is, just how long has that tradition existed? and, are we talking about a continuous, uninterrupted existence, or something else?
In traditional cultures, one of the principal vehicles for preserving and transmitting tradition is oral history. There are others too, of course, such as symbols, dance, chants, songs, and even material remains like architecture and tools. But, I am not going to speak to these. I am only going to focus on oral history for now; namely because it was one of the core elements of our field season in Kosrae this year.
But before I continue, I want to say a few words about oral history. It is a rich store of cultural lore handed down from one generation to the next; it is rooted in the past, in the 'before the before,' and is especially important in those regions of the world where current generations retain long, direct and uninterrupted lines of descent. For Kosrae, oral history presents a limited avenue of inquiry-as the arrival of missionaries in 1852 brought almost complete suppression of traditional knowledge. Where it does exist, oral history is used to explain events of the past, to preserve family histories and relationships (one of the most important features as genealogies form the backbone of oral history); it is also used to chronicle issues of protocol and proprietorship, to reinforce belief systems and moral codes, to describe traditional activities including the expressive and industrial arts, to entertain, and even to breath life into the archaeological text. I've heard it said that oral histories are for people who make it their business to remember the affairs of the community; mainly, they remember the things that interest them the most, like marriages, deaths and children.
As in oral histories in other places, time is relative. Temporal depth is difficult to identify and pin-point, which makes it more difficult to assign a story or event to a specific time period. This, unfortunately, is a natural by-product of oral history-stories passed from one generation to another, with variations arising with each retelling, usually under new circumstances, in different environments, and often under the influence of varying life circumstances and histories of the tellers themselves. But, the very essence of a story tends to remain immutable; only the descriptive text that surrounds this core and propels the story forward, is subject to change in many subtle ways.
This year, I was asked by the Kosrae Office of History and Culture Preservation to direct a reconnaissance survey and mapping program in the Upper Tofol River drainage on Kosrae. We had worked there before, about two years ago, and excavated a small site that produced a series of dates beginning about A.D. 1100 to 1200-a time period that marks the onset of an era of independent polities, each fighting with one another for dominance over the island as a whole. This is also an area that is awash in traditional stories of internecine conflicts, power struggles, love, revenge, and rebellion against despotic rulers.
Our task was very simple-we were to:
(1) penetrate the island interior (well, the margins of that interior) and find out what kind of sites are present, using the drainage basin as our geographic parameter;
(2) follow out the stone path we identified during the previous excavation work (this was something new for the archaeological record on the island, as up until our work the standard archaeological settlement model was one of dispersed isolated homesteads scattered across the coastal plain, with no sites located on ridges, ridge slopes or in the island interior, with the exception of Menke. The stone path suggested something else-a network of interconnected sites and routes that could potentially cross the island's interior, bringing further reality to the ability of Sinlaka, the principal deity, and her priestly minions to cross the island 'in the blink of an eye'); and
(3) find the one site that is known to every school child on the island, the site of Finol Tokosra, the King's Mound-where the King of Leluh was killed in a brutal act of revenge. Throughout our work, the story of Finol Tokosra provided a kind of undercurrent. It was always there, but we did not let it direct or influence our observations and ground interpretations; we kept it in our back pocket, so to speak, checking it now and then, comparing features in the story with those on the ground and asking ourselves all the while, is this it? or this?
The Story of Finol Tokosra
The bare-bones version of the story with which we started our fieldwork is retold below. There are a few variations to this story, as you might expect; many of which we gathered over the course of our time in the field. These variations are outlined following the story of Finol Tokosra.
One day a woman was fishing along the shore. She saw the king and his men coming near her, so she wrapped her fish in a mat. She did not want to give her fish to him, but in those days, if a king [chief] asked for something, you were obliged to provide it, or suffer the penalties, which was usually death.
When the king and his men arrived, the fish were still alive and wiggling, making the mat move around in very odd ways. The lady told the king her baby was wrapped in the mat. But soon, the fish broke through the mat and revealed themselves.
Angry at having been lied to, the king cut off her head.
Her family, who lived in Tofol, wanted revenge. The people of Tofol wanted to get rid of this king too (he was particularly demanding of them, and they were tired of fulfilling all his requests), so they were in full support of the plot for revenge. They decided to invite the king and his men to a large celebratory feast. Then, they proceeded to dig a very large hole and gather food for the feast.
On the appointed day, they called for the king and his men. When they arrived, all the people of Tofol were already celebrating. They picked up the canoes with the king and his men still in them, hoisted them above their heads and passed them to the front of the crowd. There, the canoes, the king and his men were thrown into the big hole and every one from Tofol stoned them to death, filling the hole completely with stones. Then, they had their feast.
As our fieldwork progressed, more and more variations to the story came forth. In these versions:
* Tofol had a very large population of some 10,000s of people; some say up to 30,000 people. When they gathered to greet the king and his men upon arrival for the feast, they lined the Tofol River, from the shoreline to the site. The crowd was thick and in a festive mood. They picked up the canoes with the king and his men still inside, passed them overhead, and then threw them into a hole and stoned them.
* When the king and his men were killed, every one fled Tofol to the southern end of the island-to Walung and Malem-in fear of a war of reprisal with the Leluh people. At that end of the island, we have stories about the sudden arrival of refugees running from the king.
* At the time of this story, there were no mangroves, only a sandy shoreline. When people had to carry anything from the lagoon, they lined the Tofol River from the shoreline to the site, and passed the item from one person to the next. This is how coral came to be deposited on the site.
* The woman killed by the king and his men had two grown sons. It is they who planned the revenge of their mother's death, assisted by the people of Tofol.
According to this variation, when the king and his men arrived for the feast, throngs of people from Tofol had lined the Tofol River from the shoreline to the site. They hoisted the two canoes with the king and his men inside over their heads, and passed them toward the front of the crowd until they reached the top of the mountain. There, they threw the two canoes and their occupants into a large hole and buried them in rocks.
Today, the ends of the canoes can still be seen sticking out of the mountain; these are the breasts of the island's Sleeping Lady profile. The two brothers killed themselves after having caused the death of the king and his men.
* There is a path along the Tofol River, from the shoreline to the site. We found portions of this path about two years ago.
Without knowing exactly where the site was located, or if it even physically existed-is it something from legend? is it a real place from which a legend grew exponentially?-the only other information we had to go on was gleaned from some of the older people on-island. Many said that the site would be large and have a coral pavement. Coral paving is a rare feature, particularly in inland sites (not adjacent to the coastline), and usually indicates a place with special status. The other bit of information we collected was from a man who used to operate a bulldozer for Public Works. He was responsible for building the access roads to a new dam on the river more than 20 years ago. When he came upon a site with elaborate, well-formed walls, he immediately stopped because this, he was sure, must be Finol Tokosra. He was so sure of the identification that he re-routed the roads to avoid the site; other construction work penetrating the site was also stopped at this time. Unfortunately, when we spoke with him, he only had a vague recollection as to where the site was located ('someplace with stone walls' that you could reach with a bulldozer).
At the end our first week in the field, we identified 26 archaeological sites within an area of about one square kilometer of dense jungle and highly dissected terrain. All of these sites are new to the archaeological record of Kosrae. Of these, we mapped one completely; we mapped a sample of another one; we mapped the Tofol River, its tributaries and a second river that feeds into the drainage basin; and we walked away from the project with new insights into site patterning on the island.
Our initial reconnaissance was both systematic and opportunistic, and resulted in the identification of a whole network of sites throughout the drainage basin, from the river courses to the ridge tops, including mid-slope benches, hanging valleys and virtually any other parcel of ground between these areas that may have accommodated a cultural presence. We identified several villages (two of which became our main contenders for Finol Tokosra, at least initially), boundary sites, defensive locations, special-use sites, transportation routes, and traveler's way-sides. Our mapping work along the rivers and their tributaries provided additional insights into site access with the appearance of stone-lines staircases, ramps and pathways that joined virtually every site to the river. Rivers, as the oral histories had already relayed, were the main travel routes from one place to the next. But, what those oral histories didn't say, was that stone paths paralleled those rivers where the banks were wide enough and flat enough to accommodate them; where there were no banks, the riverbeds themselves were the paths.
What surprised us most of all, however, was the increasingly complicated architectural details that emerged as the vegetation was removed from our selected mapping sites, our two contenders for Finol Tokosra.
This was our first contender for Finol Tokosra, and as we continued working here our confidence increased that this was probably, likely, more than likely, that legendary place. Over the course of our project, we mapped the entire site. The site extends from the Tofol River to the base of the ridge slope defining the drainage boundary; it occupies the entire river terrace in this part of the drainage basin. It is centrally located, at the point where the river course opens onto a broad flat plain that gradually descends to the shoreline, the mangrove and then the lagoon. This particular site contained an extensive feasting area with a crushed coral pavement, elaborate housing compounds of multiple chambers, a number of associated kitchen areas complete with um and faa'faa stones, a stone-lined spring with a paved catch basin (this is the only recorded site on the island with such a spring), a medicine-making area, several sakau production areas, pathways throughout the site, an elaborate drainage system to accommodate run-off from the spring, elaborate staircases connecting the upper and lower tiers of the site, and a boulder pavement that extends across the entire site.
Just outside the site, at its main entrance, there is an exterior compound complete with cooking facilities. This exterior compound is comparable to resting places observed at other sites, and was probably intended to house weary travelers before they enter into a main village (here they may take their rest and repose, and prepare themselves for formal entry and introductions). And, in a secluded part of the site, away from the main entrance, there was a very specialized feature-a sakau-sitting stone. The sakau-sitting stone has a special function, again drawn from oral history. It is used only on special occasions during the year to produce a highly potent form of this narcotic drink. This particular ceremony, according to oral history, requires the participation of a young, naked woman who must be a virgin. She is supposed to sit upon the rock with her legs spread. As the sakau is squeezed through its hibiscus mesh, it is allowed to run from the girl's breasts, down her belly and over her pubis, to then drip into a coconut shell held between her legs. According to the oral history, this produces a highly potent, very pure form of sakau.
At this site, there is one feature in particular that stands out as highly unusual-an observation point defined by oversized boulders forming the exterior corner of one compound. The boulder construction is excessive for the structural needs of the corner (the fact that no such construction appears at any other wall corner at this site, or any other for that matter, seems to reinforce this point), and is directly aligned with the two peaks on the island's Sleeping Lady profile-her two breasts described as the ends of the king's canoes in the story of Finol Tokosra. Is this mere coincidence? Possibly, but unlikely owing to the deliberate alignment and expense in terms of construction effort at this particular location in the site. Could this observation point be directed toward other observations? Such as stellar sightings, using the island's mountainous profile as a measuring device? Or could such a place be used in a different way, to perhaps recount the story of Finol Tokosra, with the backdrop of the story-teller being the two canoe peaks?
Other features across the site include the remnants of historic era disturbances, most of which are confined to the upper level of the site where the feasting area is located along with the more massive architectural compounds. One such disturbance was Japanese farming, which produced visible furrows and sweet potato mounds. Other disturbances occurred during the Trust Territory administration, namely bulldozed roads used for access to the dams on Tofol River, heavy equipment scavenging for road base stone (much of the boulder constructions on this upper level were damaged as a result), several backhoe trenches also primarily related to the road and dam construction, and some more recent experimental agriculture station farming.
This was our second contender for Finol Tokosra, based initially on its size along. This impression soon faded as we worked on the site, clearing portions which were to be part of our mapping sample. Yes, this was a large village on the river terrace opposite Site Ko-A11-32, and in the archaeological record of the island, this alone made the site both interesting and unusual. It also contained an unusual feature that piqued our interest-several stone mounds scattered about the site, each of which were near the foundations of structural compounds. None of us had ever seen such mounds-which consist of neatly stacked piles of fist-sized stones stacked between two to three boulders and covering an area with a diameter of about one meter-they were new to the archaeological record of Kosrae. These piles prompted our own speculations, and the development of several hypotheses:
1. If this site were Finol Tokosra, then the mounds could be in preparation for stoning the king and his men. But, the stones were unused; they were still in place. Were they held in reserve?
2. Alternatively, the stacks of stones could be in preparation for conflict. Internecine warfare was ubiquitous across the island, according to the oral histories, and one should always stand ready for conflict.
3. The stone mounds could be Japanese sweet potato mounds. This was ruled out, as sweet potato mounds use much larger rocks, and are never piled on boulders.
4. Neatly piled stacks of cooking rocks? Except that none of us have ever encountered such tidiness when it comes to cooking rocks; mostly they are spread outward from an um, earth oven, as the residuals from rake-outs.
5. Burials. We don't know how the dead were treated in the past, other than the chiefly elites. Drawing upon historical patterns, where burials are concentrated within a living compound, could this be the past equivalent of such a pattern?
6. Calendar rocks. A common theme in oral histories is setting up some form of a reminder for an important event, often with stones stacked on anniversaries to mark the passage of years. Was this such a place? Why then, do we not see these features at other sites?
7. Fermented breadfruit pits; a kind of traditional refrigerator to preserve famine foods. Except for the fact that such pits are only known from locations near the shoreline, where they take advantage of the tidal influx of salt water to aid in the preservation process.
8. A practice area for Nanparatak before he went to Pohnpei. Nanparatak is the hero of Kosrae, who defeated the enemies of the Pohnpei's King of Kiti by catching and then throwing stones. Alternatively, was this a site where warriors imitating Nanparatak's technique practiced?
9. Post bases. This became our most likely hypothesis, as we continued to expand our mapping area of the site. What became quite clear is that each mound of rocks was associated with wall alignments and structural foundations.
By the time we finished our mapping sample of this site, it's initial role as contender for Finol Tokosra diminished dramatically. It still remained a highly complex, very interesting village with an extensive complex of walled compounds, a medicine-making area, a network of paved drainages, several pathways connecting one feature to the next, a larger and more elaborate compound just beyond the central core of the site, several pathways connecting the village to the Tofol river, a couple of associated (potential) bathing areas within the river, and virtually no historic era disturbances.
What the site didn't have was a large public feasting area or other features that would suggest its use as a central place. Also lacking is the monumental architecture, the coral paving, the extensive basalt boulder paving, the spring, the sakau-sitting stone, and the observation point that seems to link the other site directly to landscape features that played a key role in the oral history. In fact, the more time we spent at this site, the more it seemed to be a site in-service to the previous site.
On both sites, Ko-A11-32 and -33, wall construction consisted of basalt boulders retrieved from the adjoining riverbeds. The construction itself followed traditional techniques of dry-wall masonry, with the walls raised in multiple courses and filled with randomly placed stone rubble. Wall corners, entries into walled chambers and even entries into the sites themselves were flanked by either vertical uprights or large boulders. Stone paths and pavements were composed of small boulders embedded in the soil to create leveled surfaces, with successive tiers built to accommodate and level sloped terrain. Much of the wall constructions had collapsed. In fact, in none of the sites we identified over the course of our reconnaissance did we find standing walls; wall collapse was uniformly present at all sites, unlike sites on the other side of the island, where walls remain well preserved, intact and standing tall. Why the difference? Was it related to cultural/political forces, a product of the weather, or something else? One principal difference between the Tofol drainage and other places around the island is that the Tofol river opens directly onto the eastern shoreline of the island, where the full force of ocean-borne storms slam against the island. During our project work, it rained, or rather stormed, almost continuously (and yes, we continued to work). New details within the site architecture were revealed after every storm, thanks in part to vegetation being laid flat, and the rivers of water and sediment that washed away (redistributed?) the soils packed between stones.
Our work within the drainage basin also resulted in new ideas about the patterning of sites and landscape modification, which were contrary to the former settlement model developed during the Trust Territory occupation of the region. At that time, site patterning on Kosrae was described as one of isolated homesteads dispersed across the coastal plain with no sites on ridges or ridge slopes, no villages, and no pathways. We found something different. We found a complex record of sites located on virtually every inhabitable landscape feature, with a density that covered the slope faces from the river terraces to the ridge tops. The range of site types (admittedly, our own designations) included villages, boundaries, defensive locations, special-use sites, transportation routes, and travelers' way-sites.
Most villages were located primarily on the river terraces, with at least one village located in an hanging valley. Paved pathways were observed along the river terraces, some leading into the rivers and some leading directly into the village sites. Ridge tops too seemed to fulfill several functions, with the principal function being that of settlement boundary. Just below the ridge tops, formal stacked stone walls were observed. These walls followed the length of a ridge, as it defined a specific valley system. The ridge top itself supported archaeological features that appeared to hold strategic value, such as look-out points. At the end of one ridge line, for example, just before it drops into the mangrove (and lagoon), one feature looked as though it was used to light signal fires that would have been visible from the sea. Other features along the ridge line appeared similar to house foundations, yet with no visible associated village structure. This was initially confusing, particularly with boundary walls just below the ridge line Were these guard houses? Or, was it possible that ridge lines were essentially neutral territory for travelers crossing the island? Potentially, the house foundations could be way-side stops for these travelers (that, at least, is how we are interpreting them). Mid-slope benches too, between the boundaries and travelers' way-sides on the ridge tops and the villages on the river terraces, contained smaller habitation sites complete with house foundations, kitchen areas and sakau-making areas. And, within the rivers themselves, large boulders displaying the results of generations of tool sharpening were also observed.
Overall, the pattern of settlement throughout the drainage basin was one of an interconnected network of villages and smaller sites, each strategically located in defensible locations. Boundaries were equally visible, and placed in such a way as to allow travelers to continue along their passages unprovoked. The one component missing from this observed pattern is chronology. Surface features alone cannot tell us when the drainage basin was occupied, or even if all these sites are contemporary. Retrieval of that kind of information-timelines-requires a program of excavation and the recovery of datable material from secure contexts.
A Concordance of Evidence?
So, what do we have? Are either of the two selected sites Finol Tokosra? The former site, Ko-A11-32, contains several features that make it a very likely candidate. It appears to be a central place, utilized for public gatherings. It has an expansive feasting area, set apart from other sites by the crushed coral pavement; it has several other features that support celebratory functions, such as sakau-making and faa'faa production areas. It contains elaborate, monumental architecture, with much of this located on the upper level of the site-essentially the public face of the site. Entry into the lower level of the site, to the back of the site, requires descent down one of several staircases. Also present at this site is a highly unusual feature-an observation point aligned with the twin peaks on the island's profile (the canoe ends in the story of Finol Tokosra).
The latter site, Ko-A11-33, does not contain any of these features. It is not a central place; it does not have a feasting area, nor any of the necessities any feast or celebratory function would require; it does not have any public architecture or monumental features; in short, it lacks virtually every feature found in the former site. Instead, it appears more like a supporting site, in-service to those occupying the former site.
Does the first site line up with the features in the story of Finol Tokosra? Perhaps the coral pavement mentioned; perhaps the feasting area expansive enough to accommodate large crowds, together with the necessities for celebratory functions, such as ums, sakau and faa'faa stones; perhaps the monumental architecture making the place suitable for entertaining a king; perhaps the central location along the Tofol River, within easy reach of the shoreline. Would these be definitive links between site and story? At this point, all links are coincidental; we have nothing concrete to make a solid connection. There is no timeline, for example; nor do we really know if the story itself is based on real events. But, when told within the context of the physical setting of this particular site, particularly with the backdrop of the twin peaks looming overhead, the story of Finol Tokosra comes alive. It becomes believable; perhaps that is the main function here, to dramatize a legendary story that recounts events that transpired in the 'before the before.'
The whole island is convinced we found THE site, Finol Tokosra, but I am not so sure. I continue to ask, what would the specific features of such a site really look like? Will we ever know if we identified a specific location discussed and described in the oral histories? There are plenty of examples of physical locations identified in oral histories. Usually, they are known through some form of tapu-as reference to places where special behavior is warranted, a sacred place, a place likely to invoke some terrible penalty if violated. Such places do exist, like the dwarf's mangrove, Nanparatak's spear, Menke Village (home of Sinlaka, the breadfruit goddess and chief deity on the island), Likinlulem (the place from which the island's traditions and title system originated), and Finlas Lap (the source of red mineral pigments used to paint canoes, buildings and important items). This suggests that places like Finol Tokosra should also exist, and have a physical counterpart.
But with Finol Tokosra, we are also missing some crucial pieces of information (like specific location within the Tofol drainage), which relegates any identification we make to the realm of hypothesis only. Our project work has certainly raised a number of interesting questions-some of which may be addressed through more intensive archaeological work, such as excavation. Timing and the temporal position of our site is particularly important, along with information about activities undertaken within specific areas of the site. Information that can be drawn from the full network of sites within the drainage basin include settlement chronologies, changes in architectural style over time, changes in village site plans, and use of some of the specialty sites (the boundary sites, the traveler's way-sides, even the strategically defensive positions-was there really a look-out point and place to light a communications fire at the end of one ridge line overlooking the lagoon?).
The continued collection of oral histories can help us here too. For Micronesia, they provide an important supplement to the physical record. There may be no temporal dimension that can be easily identified within a narrative, but the content and form often betray a kernel of truth derived from actual events. Such stories suggest investigative paths and alternatives in proving or disproving interpretations drawn from the archaeological record. In the case of Finol Tokosra, our work has just started. We still have a long way to go before arriving at any conclusions.
Funding for this project was provided by the U.S. National Park Service, Department of Interior, in partnership with the Kosrae Office of History and Culture Preservation.
However, the contents, interpretations and opinions expressed in these pages do not reflect the views or policies of either the Department of Interior or the Kosrae HPO-these remain solely the responsibility of the author. Also deserving of special mention are the many people who participated in the fieldwork-some as members of the core field crew, others as part of the archaeological training program in site mapping and interpretation conducted concurrent with fieldwork. The core field crew included: Hamlet Jim, Joseph Jonithin, Houver Alik, Kilafwa Mike, and Hosia J. Alokoa-they worked the full-length of the project, from the initial reconnaissance, through vegetation clearance, site mapping and interpretation, to the final site clean-up and removal of the vestiges of our various activities. During site clearing activities, and in preparation for the mapping work, a crew of five prisoners - Merle Jackson, Lynix Langu, Talley Talley, Bingham Nena, and John Phillip-and their guard, Sgt. Clarence Alokoa, assisted Dr. Beardsley, Mr. Lonno and the core archaeological crew in clearing much of the vegetation from site Ko-A11-32. Additional project participants included Mr. Lui Xinyi and Mr. Wen Rui (Sum), both archaeology graduate students from the University of Science and Technology of China; Ms. Josepha Maddison-Hill from the Marshall Islands; Mr. Jose (Joe) U. Garrido from Guam; Mr. John Castro, Mr. Herman Tudela, Mr. Vincent Pangelinan, and Mr. Lufo Babauta from Saipan; and Mr. Bruce Brandt from Kosrae, who also took many of the more useful photographs of the sites.