View more image of Ancient CHamoru Pottery here.

What artifacts tell us

Archeologists help us learn about ancient cultures by looking at the objects or materials, known as artifacts, that people left behind. The most abundant artifact found in archeological excavations throughout the Mariana Islands is pottery. Literally thousands of potsherds (fragments of pottery), strewn across different excavation sites, have been collected and analyzed over the last few decades. What archeologists have found is that CHamoru pottery not only came in a variety of shapes and decorative motifs, but that pottery styles underwent significant changes over time.

These changes possibly reflect other changes in CHamoru lifestyle patterns, including settlement into more permanent villages or other social and cultural factors.The earliest Pre-Latte Period pots found in the Marianas were generally wide-mouthed, shallow and decorated with lime-incised designs, while vessels from the Latte Period were larger and heavier pieces with few decorations. In any case, the clay pot was one of the most versatile objects used by the ancient CHamorus. It could be used to store or cook food, boil water, or hold or transport other items. Pots have also been found within a small number of human burials.

Although not much is known about the specific techniques used by the CHamorus, pots in the Marianas most likely were manufactured locally by clay from deposits found around the islands. Using hand-building methods, whereby an individual pot is formed from coils of clay stacked upon each other, or balls of clay molded into the desired form, these vessels were then fired to harden the clay and give it strength. The clay were mixed with temper inclusions of coral or volcanic sand before firing.

The CHamorus did not use potter’s wheels, nor did they have large ovens or kilns used for firing clay. Instead, they may have used open fire or pit-firing methods to produce their ceramics until the introduction of the kiln by the Spanish. Unfortunately, traditional pot-making methods have been lost since the era of Spanish colonization of the Marianas. Although an unbroken pot has never been recovered, some archaeologists have been able to use large fragments to reconstruct a number of vessels and piece together a history of the ancient CHamoru through their pottery.

Ancient CHamoru pottery

Fragments of ancient clay pots, known as pottery sherds (or potsherds), are widely distributed across Guam.  They are particularly abundant on the surface of Latte Period villages (from about 500-1,000 ago), and especially in those villages located on the coast with intact above ground features, such as latte sets, lusong (stone mortars), wells, and midden (refuse) mounds.  They are less common on the surface of interior latte villages, and although they are there in smaller numbers, they are found in rock shelters, and in places used as temporary encampments.  The sherds form surface accumulations called pottery scatters and pot drops (a place where one pot broke).   The sherds also are always found in excavations at buried archeological cultural deposits where they are often associated with cooking features like hearths or earth-ovens.  Over the years, tens of thousands of pot fragments have been recovered and thousands have been analyzed.  What has been learned about the people who made and used the pots?

The distribution of pot fragments (a whole pot has never been found) across the island indicates that the pots were important to people and were used in many places and probably in many ways, such as for cooking, serving, displaying, storing, and perhaps transporting some foods or liquids.  Over time the shape, size, and the ways that the pots were finished, as well as the types of inclusions added to the clays during pot manufacture, changed.  Some of these changes reflect stylistic preferences and others may be related to function.

For example, the shape, size, finish, and temper inclusions seen in the Latte Period pots differ from those made earlier in the Mariana’s ceramic sequence.  Archeologists propose that these variations in form may be related to function.  If true, then the new forms may indicate that people made some significant changes in their cooking and storing requirements near the beginning of the Latte Period.  The reasons for these changes remain open to discussion.  It may have to do with an increase in population size, the introduction of new crops or farming techniques, or other social and/or climatic factors.

Latte Period pottery forms

It seems the rounded vessel shape, volcanic temper inclusions, crudely formed walls, and the incurved rims of the Latte Period pots formed appropriate containers for boiling foods like taro, yams, breadfruit, rice, preparing soups, making salt, making coconut oil, and making sugar or syrup from sugar cane. The analysis of the charred residue on Latte Period sherds has identified starch from taro (Colocasia esculenta), indicating that these tubers were one of the foods cooked in the pots. Also identified were raphides (needle-shaped crystals) from Colocasia and Cytrosperma taro plants, which indicate that these leaves had been used. The leaves could have covered the food while it cooked, or been prepared as one of the dishes to be served. Accidental rice impressions identified on other Latte Period sherds indicate that this grain was available to the people; it may have been cooked in the pots, but as yet, no rice starch has been identified in the residue. Starch grains and raphides from the ti plant (Cordyline) on other sherds suggest that this sugar-rich tuber was cooked and its leaves used. Angular, fractured bits of marine shell on another sherd suggest that shellfish were prepared, kept, or served in these pots.

Interestingly, the charred residue occurs on sherds with different surface treatment categories, including plain, wiped, punctate, and faintly combed.

The strong association between surface treatment and function has yet to be established for the Latte Period pots, though it appears such pots were used for cooking.

The large capacity of some of the Latte Period pots suggests that large quantities of foods were prepared. Some large pots could have been used for food and water storage. The grooves on some fragments suggest that a cord network was used to suspend the pot, secure a lid for safe storage, or provide a handle for gripping or pouring. During the late Latte period, a new pot with a thickened, everted rim developed and apparently spread across the island. Rims from this form have been found at sites on the east coast, west coast and in the interior of the island.

Similar surface treatments are seen in the pottery collections recovered from most Latte Period sites. Although there may be some slight differences in the percentages of the different finishing categories (plain, rough, combed, trailed, wiped), it appears that Latte Period potters attempted to achieve similar specifications when making their pots. This suggests that the people making and using the pottery shared the same culture.  Apparently the pottery was not used to convey information about separate group affiliations within the larger social community.

Pre-Latte pottery forms

The pots made prior to AD 900 were manufactured before the Latte Period, a time known as the Pre-Latte Period. Some of the Pre-Latte pots made from about 500 BC to about AD 500 had flat-bottoms with short vertical side walls. Pots with this shape are not suitable for boiling foods. However, fragments of these pots are associated with areas of fire altered rocks and charcoal, which suggests they were used for a different kind of cooking, such as frying, roasting or steaming. Breadfruit and/or flat cakes made of the flour from gabgab tubers, or some other plant, may have been prepared in these pots.  The analysis of the remains of charred residue on a few sherds dating to this interval identified starch from taro and Cordyline.  A fish scale fragment and fractured bits of marine shell suggest these pots were also utilized in the processing of fish and shellfish.

Some of the flat bottomed pans were quite large; they would have been heavy and not easily moved from place to place. Archeologists have proposed that the appearance of these large pots around 500 BC indicates that people became more settled then. Generally, the flat bottomed pans are found in coastal villages and in the inland portions of these villages.  Flat-bottomed pans are rarely found in archeological sites located in Guam’s interior. This is probably due to the fact that most interior sites were utilized after AD 500, and by then, the pans had dropped from the ceramic sequence, perhaps no longer manufactured or used.

The shape of the Pre-Latte pots made prior to 500 BC differs from the pots described above. Very little is known about how these early pots were used.  Some have a globular shape with a restricted mouth and a recurved, unthickened rim, calcareous (chalky or made of calcium carbonite) sand temper, and a red slip finish. These may have served as cooking pots.  Others have an open mouth with a flaring sidewall and a rounded bottom.  The open pots could have been used for display or for ceremonial events. A small number of the open pots had been decorated with lime-filled designs (impressions and incisions).

Over time the designs evolved from complex to simple patterns, and by AD 500 the decorated pots entirely drop from the ceramic sequence.

Since sherds with similar impressed and incised designs have been recovered from the early sites located in the four largest islands of the Marianas, some archeologists think that the people of this time had a mobile lifestyle, as the smaller pot sizes would have been easier to transport. Perhaps they utilized the resources in one area for a while and then moved to another area. Such a settlement system could indicate fairly low population numbers and the islanders’ continued ability to cross the open ocean distances that separate the islands from one another.

Studies of the clay in the sherds found that different pots from a single site had been made from different clay sources on Guam. The use of more than one clay source has been interpreted to mean that pots could have been made in different places by different people. It also means that people carried clays or finished pots from one place to another. As yet, little is known about the social factors involved in the clay/pot movements.  Historic accounts of the traditional pottery making industry have not been located, and it is not known exactly where pots were made, or who made them.  Archeologists have identified a few sites where pottery was probably made and/or fired on Guam, but the clay studies suggest there may have been many other places.

Inter-/Intra-island pottery exchange

Temper inclusions have been studied to gain information about inter-/intra-island pottery exchange systems. For example, fairly large and prominent inclusions of quartz have been noted in some sherds. Quartz-rich deposits are differentially distributed across the Mariana Islands. These deposits occur in Saipan, but not in Tinian, Rota, and Guam. Large quartz inclusions are commonly seen in Saipan’s pottery. Since fewer quartz-tempered sherds occur in Guam, it appears that a systematic exchange of pottery between Saipan and Guam was not in place during ancient times. However, the pottery was moved around. Sherds made on Guam have been identified on Aguiguan, Tinian, Rota, and Saipan. Sherds made on Saipan have been identified on Tinian, Rota, and Guam. This distribution indicates that the people from the different islands in the Marianas continued to interact with each other from earliest times to European Contact.

On Guam, early pottery has been recovered from sites located along the coast.  The sites with early pottery and early radiocarbon dates include Mangilao Golf Course (the Mochom site), Tumon Bay (Ypao, Tumon, Naton), Ritidian Beach, Tarague Beach, and Hagåtña (Plaza de España and behind the Cathedral). Generally, the buried cultural deposits containing the earliest pottery are located on the furthest inland portions of the sand flats at these sites. The beaches were narrower 3,000-4,000 years ago because sea level was higher than it is now. The sites listed above have yielded pottery associated with radiocarbon dates ranging from about 1,500-1,000 BC. Tumon Bay on the west-central coast of Guam has provided earlier radiocarbon dates, but they are not associated with pottery.  Likewise, the oldest radiocarbon dates from charcoal bits in cores taken from Guam’s wetlands are not associated with pottery.  Based on the current pottery associated dates, it is not possible to determine exactly where people first landed on Guam.

By Darlene Moore, MA

Word list

Temper – In ceramics or pottery-making, “temper” refers to the material (or inclusions) that are mixed with clay to improve its workability or give it other qualities. Without temper, the clay may not fire properly or break easily during firing. Tempers give the resulting ceramic different finishes or surface textures. In Marianas pottery, he observation of tempers can indicate which era the pottery belongs. A whitish, coralline sand temper often indicates Pre-Latte pottery, while darker, volcanic sand tempers are representative of Latte Period ceramics.

Pottery Slip – is formed by the addition of water or a smooth layer of wet clay to the surface of a pot before it is fired in order to produce a smooth finish.

Surface Treatments – refers to the kind of texture that is present on the surface of a pot. Punctate treatments resemble indentations, spaced fairly regularly, to produce a design. Such surface treatments were applied by hand or with other tools, even fibers from coconut husks. Although the designs on pots from the Latte period are less decorative and more functional, it is suggested that the designs provided a surface to easily grip or carry the vessel, as CHamoru pots had no handles.

For further reading

Athens, J. Stephen, Michael F. Dega, and Jerome V. Ward. “Austronesian Colonisation of the Mariana Islands: The Palaeoenvironmental Evidence.” Bulletin of the Indo-Pacific Prehistory Association 24, (2004): 21-30.

Butler, Brian M. “Pots As Tools: The Marianas Case.”  Micronesica Supp. 2, (1990): 33-46.

Commonwealth of the Northern Mariana Islands Division of Historic Preservation. An Archaeological Survey of Aguiguan (Aguijan), Northern Mariana Islands. By Brian M. Butler. Micronesian Archaeological Survey No. 29.  Saipan: CNMIHPO, 1992.

Dickinson, William R., Brian M. Butler, Darlene R. Moore, and Marilyn Swift. “Geologic Sources and Geographic Distribution of Sand Tempers in Prehistoric Potsherds from the Mariana Islands.” Geoarchaeology: An International Journal 16, no. 8 (2001): 827-854.

Graves, Michael W., Terry L. Hunt, and Darlene R. Moore. “Ceramic Production in the Mariana Islands: Explaining Change and Diversity in Prehistoric Interaction and Exchange.” Asian Perspectives 29, no. 2 (1990): 211-233.

Moore, Darlene R.  “Measuring Change in Marianas Pottery: The Sequence of Pottery Production at Tarague, Guam.” MA thesis, University of Guam, 1983.

Pago Bay Resort, L.L.C. Archaeological Research at the Laguna Pago Bay Resort, Lots 155 NEW, 164 NEW, and 164-4, Yona Municipality, Guam. By Mike T. Carson and John A. Peterson. Mangilao: MARC, 2009.

Paul H. Rosendahl, PhD, Inc. Volume II: Data Analyses, Archaeological Mitigation Program, Mangilao Golf Course Project Area. By B.J. Dilli, A.E. Haun, S.T. Goodfellow, and B. DeRoo. Hilo: PHRI, 1998.

Reinman, Fred. An Archaeological Survey and Preliminary Test Excavations on the Island of Guam, Mariana Islands, 1965-1966. Mangilao: Micronesian Area Research Center, University of Guam, 1977.

US Department of the Navy Naval Facilities Engineering Command, Pacific. Guam’s Prehistoric Pottery and its Chronological Sequence. By Darlene R. Moore. Honolulu: IARII, 2002.