Mauna Kea
Public television documentary
News updates
Additional resources

July 1977
Patrick McCoy/Richard Gould

On the upper slopes of Mauna Kea, it hardly seems like Hawaii. Certainly it looks nothing like the islands of the travel brochures, for there are no trees above the 10,500 foot elevation on this great extinct volcano. The few plants are mainly small ferns and alpine-adapted species growing furtively between bare rocks or in crevices. From November until May, snow rests heavily over most of the landscape at this elevation. Even in summer, storms bring thick clouds and fog as well as strong winds and hail. The air is thin, leaving unacclimated visitors gasping for breath and sometimes suffering more serious symptoms of altitude sickness. Amid these severe surroundings, far form sandy beaches and swaying palm trees, it is hard not to wonder what the ancient Hawaiians ever saw in the place.

Located on the island of Hawaii, Mauna Kea, elevation 13,796 feet, is the highest mountain in the Pacific. Massive accumulations of molten lava from successive volcanic eruptions created its dome-like form. This shape is interrupted by more than three hundred cinder cones of various sizes, scattered over much of the mountain’s lower slopes and summit area.

Although a neighboring volcano, 13,677-foot-high Mauna Loa, erupted as recently as July 5, 1975, thousands of years have passed since Mauna Kea has been active. The volcano’s summit shows signs of extensive glaciation down to around the 11,000-foot level. Glacially transported gravels and boulders are scattered over large areas, and even partly cover several of the large cones of red cinders, presenting a stark landscape created by the active combination of past fire and ice. One of the more startling phenomenon created by these geological events is Lake Waiau, lying at an elevation of 13,020 feet in a glacially-scoured cinder cone.

In spite of its bleak appearance, this area was visited often by prehistoric Hawaiian craftsmen as a series of archaeological surveys and excavations recently carried out by the Bernice P. Bishop Museum of Honolulu under the direction of Patrick C. McCoy, has shown.

Hundreds of years before Captain Cook visited the islands in 1778, people were quarrying and flaking fine-grained basaltic rock on the volcano's south flank as a first step in the manufacture of ground and polished stone adzes. Adzes of stone and/or shell were ubiquitous items in toolkits all over ancient Polynesia and much of the rest of the Pacific. The implements were attached to wooden handles and used primarily for shaping wooden objects like canoes, houseposts and images placed on religious monuments.

Stone adzes of a distinctive elbow shape and quadrangular cross-section have been found in habitation sites on every major island in Hawaii. For many years, archaeologists have studied adzes of all forms with the aim of establishing temporal changes and the historical relationships between Hawaiian types and those from other island groups in the Pacific. Similar attention has been given to the forms of other implements, like shell and bone fishhooks. While typological studies continue and can be expected to provide useful information about prehistoric human settlement and migration in the Pacific island groups, the aims of the Mauna Kea adze quarry project are quite different.

The survey, excavations and analysis of the quarry undertaken by the Bishop Museum represent the first large scale example of a new direction in Pacific archaeology — the technological study of an important class of stone tools. Instead of concentrating on the question of the origin of the adze type, this new research is collecting information about the process of obtaining the raw material and of the manufacture. What kinds of technical knowledge did the prehistoric Hawaiians need to make these tools? What sort of advance planning was required for sustained visits to the quarry? And how did the methods of working stone reflected at Mauna Kea resemble or differ from those used in other parts of the world? In short, we wanted to begin to understand more about the human behavior behind the manufacture of this one type of tool.

The authors had been interested in understanding such process and the Mauna Kea adze quarry seemed to be an ideal place to answer these questions. Other adze quarries are known elsewhere in Hawaii, notably on O‘ahu, Kaua‘i, Kaho‘olawe and west Moloka‘i, but the Mauna Kea quarry is the largest and best preserved in the islands and in the entire Pacific region. The Mauna Kea quarry is mentioned briefly in a 1902 account of Hawaiian stone implements by William T. Brigham, a former director of the Bishop Museum, and was assigned the name Keanakakoi, or “Cave of the Adze,” by a nineteenth-century expedition.

The quarry is made up of a number of different sites, giving evidence of a wide range of activities. Compared with large prehistoric quarry workshops in other parts of the world, it reflects an unusual degree of internal complexity. Thus, we expected to obtain good information here on manufacturing and core reduction techniques and on the relationships between manufacturing and other activities.

In order to collect significant amounts of lithic raw material, the ancient Hawaiians had to deal with some unusual problems, and different areas of the quarry reflect some of their solutions.

First, of course, they had to get to the quarry. Most of the complex, which extends over an area of about seven-and-a-half square miles and is at least twenty miles from the nearest prehistoric settlement, lies at an elevation between 1,000 and 12,400 feet. Several well preserved trails of ancient origin go up the mountain and at least one passes through the quarry. Although ascent to the quarry presented no technical difficulties the climb did require considerable stamina.Other problems were presented by the need for water, fuel, and food. The climate of the upper mountain is so dry and cold that it has been described as a stony alpine desert. No perennial streams appear above 4,500 feet. Fortunately, however, Lake Waiau, at 13,020 feet — just above the quarry complex — provides fresh water within walking distance of the quarry. Mauna Kea has a definite treeline, and the dominant mid-elevation species of tree (Sophora chrysophylla, or Mamane in Hawaiian) extends, only in isolated cases, as high as the lower parts of the complex. Therefore, much of the necessary firewood had to be carried some distance up the mountain, and prepared foods had to be transported all the way from its base. In addition to the bulk of their food, the Hawaiians had to carry warm clothing, sleeping mats and containers for collecting water. Clearly, any prehistoric expedition to the adze quarry must have involved a carefully planned visit of at least two or more weeks during the relatively short summer, when the area was not covered with snow.

The constant danger of sudden storms required some form of shelter for both overnight and daytime use. A number of natural overhangs in the quarry served this purpose. Even on clear, sunny days, some protection from the wind is advisable, so less substantial walled shelters in the open were also needed for comfortable stone chipping close to the main sources of raw material. In addition to solving these logistical problems, the Hawaiians had to know how to find and extract the best raw materials. Suitable basalt for adze making is not distributed evenly over the area. Rather, it occurs in large quantities only in particular exposures and outcrops with the greatest concentration located discontinuously along an escarpment at 12,200 to 12,400 feet. In some cases, efforts had to be made to quarry subsurface deposits of unweathered basalt.

Finally, these early workmen faced the problem of transporting the raw material home. Workshops established at the quarry economized effort by allowing the reduction of each piece of basalt to a basic preform, as close as possible to the intended shape and size of the finished adze. Thus, the amount of stone to be carried long distances could be kept to a minimum. The quality of some preform rejects indicates that the adze craftsmen were very selective in what they took away.

With these problems in mind, we were able to look more closely at the Mauna Kea adze quarry and to being to appreciate the different types of past behavior the various sites within it suggest. The quarry’s most conspicuous features are huge accumulations of stone debris. Blue-grey piles of waste flakes stand out against the more mottled shades of natural weathered rock. Certain areas contain vast heaps of preforms and cores, some too large for one person to lift, along with large boulders that show scars where flakes have been removed. Basalt flakes the size of large soup bowls are common in these localities, as are round hammerstones of hard basalt that may exceed the diameter of a large grapefruit. In some cases, these stones are found close to pits cut as much as fifteen feet into the bedrock. The ancient Hawaiians extracted the best available basalt by using levers and breaking up large slabs of rock that had already been cracked by thermal expansion and contraction. Undoubtedly much of the preliminary shaping of preforms took place in these localities so our designation of them as extraction areas…is meant to reflect only the primary activity.

Even more dramatic are large heaps of smaller flakes, rejected or partly finished adze preforms, and small hammerstones. Some of these piles are one meter high and others may be as much as three meters. The largest of these always occur at the entrances to habitation rock shelters. These mounds have been tentatively labeled fine-finishing areas. (Some smaller piles of this kind occur within or very close to large extraction areas.) The fine-finishing areas appear to be places where partly reduced adze blanks were further reduced and shaped in more sheltered surroundings.

There are also numerous small work areas in which a few stone cores — sometimes only one — were reduced, leaving behind a small but concentrated scatter of waste flakes and exhausted cores. Most of these small workshops were dispersed over a wide area extending between the large extraction and fine-finishing areas in the upper quarry area, at elevations between 1,000 and 12,400 feet, and at lower elevations where scattered glacial boulders were reduced to adze blanks and then preforms. Altogether, 266 major workshop areas were mapped during the 1975 and 1976 field seasons.

In addition, we located forty small structures that we have interpreted as shrines on the basis of their similarities to religious structures common throughout eastern Polynesia. Although they vary in size and shape, their placement seems constant — they are usually located on top of rock outcrops overlooking major workshop areas. All of these structures contain rows of upright rock slabs, in some cases enclosing or partly enclosing a small, flat area. In at least one instance, a chunk of coral had been placed beneath one of the uprights; a lock of human hair was found at the base of another. Hammerstones, waste flakes, rejected adze blanks and preforms frequently lie in and around the structure, indicating that some stoneworking took place at the shrines. This may be evidence of ritual behavior intended to predict or influence the flaking qualities of basalt extracted from various parts of the quarry. Historic accounts indicate that the Hawaiians had rituals for most technical enterprises, such as planting, canoe building and fishing. It would not be surprising, therefore, to discover evidence of ritual activities related to stoneworking.

Experiments in producing replicas of Hawaiian adze preforms showed us, among other things, that many pieces of basalt are subject to a phenomenon known to stone tool makers as “end shock.” This situation occurs when a blow directed at one side of the preform dislodges a chunk from the opposite side; it results from slight imperfections that interrupt the normal elasticity of the stone. Because these flaws are not always visible on the outer surface of the rock, such breakage may have seemed unpredictable to the early stone-workers and beyond the control of even the most skilled craftsmen. Many of the rejected adze preforms at the shrines are easily identified as end-shock fragments. Because of this fact, we believe that rituals may have been performed at these structures to minimize the chance of end shock happening or to determine, through a kind of oracle, whether or not the phenomenon would occur with a particular batch of raw material.

Seventeen rockshelters were found at the quarry. All contain both abundant flaking debris and midden material indicative of intermittent, short-term habitation. Major archaeological excavations of these stratified deposits were carried out in two of the shelters, and two others were tested. Because these retreats are relatively dry and lie at an elevation with an annual precipitation of 10 to 15 inches, the preservation of organic materials is unusually good for a tropical area, where organic remains tend to decompose both rapidly and completely.

Artifacts of wood and other fibrous materials have been found in the rockshelters, among them a wood stopper for a bottle gourd, portions of wooden fire ploughs (a Hawaiian device for lighting fires), fragments of tapa cloth, braided sennit cordage, pieces of pandanus matting, and braided netting that appears to be part of a ti-lear rain cape. Numerous hearths, some of them lined with vertical rock slabs, furnished charcoal for radiocarbon analysis. Although several important samples remain to be processed, the dates obtained range from approximately A.D. 1264 to 1657; that is the mid to late part of the prehistoric Hawaiian cultural sequence.

Well preserved food remains were also discovered in the rockshelters. A detailed analysis of the faunal and plant remains is not yet completed; we know, however, that saltwater limpets (Cellana argentata), a modern-day Hawaiian delicacy, sea urchins, various fish, birds — mainly dark-rumped petrel (Pterodroma phaeopygia sanwichensis), flightless Hawaiian rail, duck, goose, chicken and small forest birds — and pigs were featured in the early Hawaiians’ diet. Remains of coconuts, gourds and kukui nuts (Aleurites moluccana, used for lighting oil) were also found. Except for several of the birds, none of these plant and animal species is found in the vicinity of the quarry. Their variety and abundance indicate the extent of advance preparations required to exploit this isolated locale.

Some of the rockshelters show evidence of alterations intended to make them as comfortable as possible. Except for two at lower elevations, all have well-constructed rock walls at their entrances. In one case, the wall consisted of slabs of rock laid in horizontal courses with others set at an angle across the top to form a roof, complete with circular smoke hole. The interior wall of this shelter had been artificially enlarged and lined with slabs. Whether improved or not, rockshelters were the best protection these people had against cold weather, and indeed were used on occasion by members of the 1975 and 1976 field parties.

Adze blanks, preforms, hammerstones and waste flakes were found in large numbers throughout the stratified deposits in the shelters. The excavated fill within the rockshelters also contained small pieces of basaltic glass — a rather poor variety of obsidian. Although their exact use is unknown, their presence with other habitational remains indicates that they were tools of some kind. Such small, thin sharp-edged flakes would work best as implement for cutting fibrous materials and wood and may well have served a variety of domestic tasks, such as shaping fire sticks or cutting strips of pandanus to repair worn sleeping mats. The source for this glass is in the main quarry area, close to the largest pile of basalt waste flakes in the entire quarry. Here we found traces of a workshop on a small, natural terrace covered with basaltic glass flakes, cores and other debris. Clearly the ancient stone chippers made whatever use they could of resources naturally available in the quarry area.

Although most of the activities represented at the Mauna Kea adze quarry — even the shrines — reflect behavior directed toward the extraction and reduction of basalt for producing adze preforms and the living arrangements that accompanied this work, other, more enigmatic features of the site are not so easily explained. At one overhang, below the rim of a large crater at the center of the quarry complex, linear, red-painted designs cover some of the smoother rock surfaces. Although they are generally well-preserved, it is hard to distinguish motifs. One design appears to represent a shark, another an octopus — both oddities at 12,300 feet above sea level. Rock paintings are known elsewhere in Hawaii, but are rare. It is more common to find designs pecked or engraved into the rock; such petroglyphs occur in a single location at the west end of the adze quarry, near a series of large extraction areas. Here the motifs include men depicted in a style seen widely throughout the Hawaiian Islands. It is not clear why these people produced these rather modest designs in this one spot and nowhere else at the quarry.

Scholars like Don Crabtree of Kimberly, Idaho, known for his work in stone toolmaking, have pointed out that the process of finishing stone tools often obliterates previous steps, making reconstruction of the manufacturing process difficult if not impossible. Once they were shaped, Hawaiian adzes, for instance, were ground and polished, sometimes over their entire surface, thus erasing most of the scars of flakes removed during manufacture. In order to understand how such a tool was made, then, it is necessary to examine not just the finished artifact but also partly finished and rejected tools, waste flakes, hammerstones and other debris. Quarries and workshops provide examples of each stage in the process of lithic reduction, from the natural stone block through the removal of flakes to the final shaping of the adze.

It addition to survey and excavation, our expedition is attempting to discover, by modern experimentation, how the ancient adze makers at Mauna Kea solved the technical problems of adze production….Like archaeological excavation, on-site experiments destroy the material they use. When an experiment simply could not be performed successfully away from the adze quarry, we made every effort to minimize the damage or alteration done to the quarry. Moreover, changes in the appearance of rocks were noted carefully so as not to create confusion for visitors or future archaeologists….

Both replicative and mechanical experiments were performed on the Mauna Kea basalt. One on-site experiment during the 1975 field season, for instance, reduced large blocks of stone, which were too heavy to be moved, to cores resembling those found at the site. A hard basalt hammerstone weighing about 35 pounds was used to remove the maximum number of large flakes suitable for making adzes. Both usable flakes — which were later made into preforms similar to those at the quarry — and waste flakes were collected and counted. On an overall average, one usable flake was produced for every twenty-two waste ones. This figure does not necessarily indicate the ratio of usable to waste flakes produced by the ancient Hawaiians, but it does provide a basis for calculating roughly the number of usable blanks they produced from the number and characteristics of the waste flakes found in the mapped workshops. Other experiments include working with hammerstones of different shapes and hardnesses; mapping the spatial distribution of waste flakes; counting, weighing and calculating the volumes of waste flakes removed in replicating adze preforms; and testing the edgeholding properties of the Mauna Kea basalt.

The picture emerging from the experiments is one of a lithic technology that differed appreciably, in certain respects, from stone toolmaking techniques widely known in both the Old and New World…

The flaking qualities of the Mauna Kea basalt and its abundance probably made this remote locality attractive for long-term exploitation. The overwhelming size of the quarry, in comparison to other adze quarries in the islands, suggests that adzes from here may have been widely traded. The research is exploring this possibility by means of petrographic analysis of basalt from Mauna Kea and other known adze quarries. If samples are distinctly different in mineral composition or amount of radioactive elements, it will be possible to determine the origin of excavated adzes in habitation sites throughout the Hawaiian Islands and perhaps beyond. Whether the distinctive features of Mauna Kea lithic technology are part of a wider Asian or Pacific tradition of stone toolmaking remains to be seen. Perhaps these techniques evolved locally, without any historical connections outside Hawaii. Such questions can be answered only after detailed comparative studies on lithic reduction techniques in other areas.

Comparative studies of the typological characteristics of stone artifacts, like adzes, can establish the basic relationships of time and space in the prehistory of particular regions. The complexities of the Mauna Kea adze quarry, however, prompt us to urge that archaeological comparisons be extended to include technological behavior as well as form. Only then will it be possible to determine when and under what conditions different lithic reduction processes may give rise to similar forms, or how differently formed artifacts can be produced by the same or similar lithic reduction sequences. By examining the behavior represented by stone artifacts, archaeologists can gain a clearer and less ambiguous understanding of historical relationships among toolmaking traditions and, at the same time, reach a more informed appreciation of the skills and problem-solving abilities of ancient craftsmen.

| P.O. Box 29 Na'alehu Hawai'i 96772-0029 | Ph: (808) 929-9659 | | Website: Still Point Press & Design Studio