War in the Pacific
Archelogy and History of Guam
NPS Logo


3. Pre-Spanish People and Culture of the Marianas Islands

At an early date, the Marianas Islands (Guam, Rota, Tinian, Saipan) were reached by people from an undetermined source, perhaps from the Philippines or by way of Palau, traveling by canoe and bringing common Pacific food plants (taro, yams, breadfruit) but no domestic animals (except, very probably, poultry; not even dogs, apparently). They possessed polished stone implements and knowledge of pottery-making.

The culture of the pre-Spanish Chamorros is known principally from early historical documentary sources of the sixteenth and seventeenth centuries (great reduction of the population and marked modification of culture took place only toward 1700), and from archeological data. These sources have been discussed concisely by Laura Thompson (The Native Culture of the Marianas Islands, Bishop Museum Bulletin 185, Honolulu, 1945) and summarized very briefly in archeological reports and general historical or geographical studies. Comparisons have been made to Indonesia particularly (see especially H. Otley Beyers, Philippine and East Asian Archaeology and Its Relation to the Origin of the Pacific Islands Population, Bulletin 29 of the National Research Council of the Philippines, University of the Philippines, Quezon City, P.I., 1948).

Little if any change in culture has been detected in the archeological remains, which appear to represent a single period of continuous occupation, and there is no evidence to suggest that this occupation was preceded by any simpler ancient culture or more primitive people (In addition to Thompson, 1945, see Laura Thompson, Archaeology of the Marianas Islands, Bishop Museum Bulletin 100, Honolulu, 1932; Douglas Osborne, "Chamorro Archaeology," unpublished ms. on Guam sites, University of Washington, 1947 — cited hereinafter as "Osborne, ms." — summarized in his "Archaeology on Guam: a progress report," American Anthropologist 49:5l8-524, No. 4, July-September 1947 — cited as "Osborne, 1947").

Recent work on Saipan and Tinian and Rota by Alexander Spoehr of the Chicago Natural History Museum, the first thorough and systematic archeological investigations in the Marianas, may profoundly modify hitherto current conceptions of the prehistory of that area. On the basis of one as yet unpublished item of new information from Dr. Spoehr — a radiocarbon date obtained by Dr. W. F. Libby on material from a site on Saipan of 127 B.C. plus or minus 200 — I have already deleted several statements in this report as to the supposed late date of immigration and brief period of cultural development. The date is associated with red-slipped pottery suggestive of Philippine ware and superior rather than inferior to the coarse Marianas Plainware of the latte sites.

The people who occupied the Marianas Islands in the sixteenth and seventeenth centuries, the original Chamorros, were brown-skinned and dark-haired, and included some who were very large — tall and heavy, with big skulls and powerful jaws — according to historical documents and judging from bones recovered in archeological sites. [1]

The language has been assigned to the Austronesian or Malayo-Polynesian stock, with especially close Indonesian connections (see, for example, W. E. Safford, "The Chamorro language of Guam," American Anthropologist V, 1903; H. Costenoble, "Chamorro language notes," Guam Recorder, XII, 1936).

The comparatively small amount of linguistic work done has, however, been apparently based entirely on the modern Chamorro language of about 50 years ago; a great deal of Filipino influence from the early 1700's on may have considerably modified a tongue perhaps originally Polynesian.

As early as 1817 it was noted by a Russian visitor that the "Chamori or Mariana language has almost vanished with the people who spoke it * * * they count only in Spanish and it was difficult to get the numerals of the Mariana language. On the other hand it appears that appellations from the Philippine language have been given to many animals and objects introduced" (Chamisso in Kotzebue 1821). A short manuscript on the Chamorro language, Lingua Mariana, by Father Sanvitores himself, is said to be in the archives of the Society of Jesus in Rome (it is hoped that a copy of this basic source will be obtained through, or at least by, Dr. R. W. Clopton, head of the Education Department of the University of Hawaii).

The population of the southern Marianas in the sixteenth-seventeenth centuries has been estimated at from 40,000 to 100,000; of Guam alone, from 3,000 to 50,000 (see discussion below, under History, Note 10). This large population lived scattered in small local settlements; villages of 50 to 150 single-family huts along the coast, hamlets of 5 to 20 houses in the interior, an estimated total of about 180 settlements in 1668 (from the abridged translation of Garcia's Life of Sanvitores in the Guam Recorder, 1937). There were over 200 houses in aboriginal Agana, according to Father Sanvitores.

The latte stones represent the sites of such villages, coastal and — in the Fena (Talafofo) River drainage — interior; there were a great many of them 25 years ago (Hornbostel estimates 270 monument sites on the island of Guam, but this astonishingly high figure might be intended for individual house-site groups, although it wouldn't be high enough to mean the separate stones).

There are only a few stone-supported structures indicated at each site, however (2 to 12 sets of 4 to 10 pairs of latte); presumably these were either the houses of chiefs' families, or else men's clubs, or both. In the description of Tinian in 1742, George Anson says the natives assured him that the many great stone pillars on that island "were the foundations of particular buildings set apart for those Indians only, who had engaged in some religious vow" (Walter, ed., 5th edition, London, 1749, p. 312).

In one historical source, the narrative of the Lopez de Legazpi expedition of 1565, it is stated that canoe sheds (as well as dwellings) were supported on stone pillars. The latte seen on Guam and others described by previous investigators do not seem high enough or widely spaced enough to shelter the praos 25 to 40 feet long, with an outrigger giving a total width of probably 15 feet or so, and with a mast and lateen sail, even if the mast were unstepped and the outrigger unshipped each time the canoe was stored. I do not, in short, find the boathouse explanation likely; on superficial consideration, it seems impracticable.

Actually, though, the idea of setting frame-and-thatch houses atop hemispherical stones precariously balanced on vertical slabs certainly appears, in a typhoon- and earthquake-ridden area, perfectly absurd offhand — but was clarified for me rather convincingly by Mr. Bert Bronson of Tumoning, Guam, to whom it occurred, in discussion of this on February 4, 1952, that lashing the frail but flexible house frameworks to the stones with coconut fiber ropes would give comparative security, with a strong though yielding resistance, in both tremors and, especially, high winds. This would, I believe, be true especially if the transverse floor members overlay the longitudinal stringers, and were tightly lashed directly to the tazas only (the heads or caps, the hemispheres), and not fastened to the haleges or uprights, permitting a little play at a sort of ball-and-socket joint in temblors.

In the Palaus in the eighteenth century, the houses "were raised about three feet from the ground, placed on large stones, which appeared as if out from the quarry, being thick and oblong; on these pedestals the foundation beams were laid, from whence sprang the upright supports of their sides" (Keate, 1758, p. 308).

The material culture of the sixteenth- and seventeenth-century Chamorros can be rather quickly summarized:

Food — various cultivated and wild plants, and fish; no meat, except seafood (and probably coconut crabs, which are relished by modern Guamanians). Crops included yams, taro, breadfruit, coconuts, bananas (by the sixteenth century, apparently — a somewhat uncertain point), rice (said in 1565 to be a major crop, presumably from China, Japan, or the Philippines — not grown elsewhere in Micronesia and Polynesia). Maize and sweet potatoes and other American crop plants were added by the last quarter of the seventeenth century. Important among edible wild plants was the federico palm, Cycas circinalis, the nuts of which were used but not the sago. The fruit of the pu-ting (Barringtonia) was no doubt used, as it still is today for poisoning fish. Betel-nut (Areca) was chewed, with lime and pepper-leaf (Piper betle) as it still is by rural older people.

Domestic animals — none, not even dogs (I know of no other area in the world to which people were not accompanied by dogs, from Australia to the Arctic, except the comparatively nearby Palaus and some of the far outlying islands of eastern Polynesia); except probably fowl, referred to more or less definitely in some of the early accounts, though at least one (the Loayza voyage as narrated in Burney, 1803) speaks only of doves kept in cages. Fowl were not eaten, though pigeons were, in the Palaus up to 1783, according to Captain Henry Wilson (Keate, 1788, pp. 300-301).

Clothing — men, none, save conical palm-leaf hats, and occasionally sandals; women, a fringe or small apron or mat on a belt. No weaving, no use of tapa (beaten paper-mulberry bark cloth).

Houses — of plants and thatch, presumably rectangular and gabled (like later and modern Micronesian houses), raised on piles, some of them on stone pillars (the latte, a unique local Marianas development), furnished and partitioned with palm-leaf mats.

Boat-houses — canoe sheds of similar construction, also sometimes raised on latte according to at least one source, which practice as discussed above, seems impractical.

Canoes — the famous "flying prao," a large, swift, and well-made sailing canoe, with single outrigger and with lateen sail of matting.

Utensils and containers — only pottery has survived to be found archeologically; very thick and very coarse hand-made (no wheel) reddish ware, not slipped or polished, with only a very little use of scratched or imprinted decoration, not highly fired and not very hard, evidently quite large wide-mouthed jars (a sample of potsherds from Guam and Rota has been brought in for examination, and will be reported separately). Pottery was not made in Micronesian islands other than the southern Marianas group and the Palaus (including Yap), nor anywhere in Polynesia, though produced in Melanesia clear out through Fiji. The southern Marianas pottery differs in certain important respects from that of Melanesia (cf. E. W. Gifford, Archaeological excavations in Fiji, Anthropological Records 13:3, University of California Press, 1951; the material from Fiji examined and discussed with Mr. Gifford at the University of California Museum of Anthropology at Berkeley January 9, 1952).

Undoubtedly baskets as well as wooden bowls were made, and undoubtedly coconut shells, bamboo joints, etc., and large shells, also were used as containers.

Implements tools, and weapons — made of stone, shell, wood, bone, etc. Boulder mortars were used for grinding, and are still to be found, rather small cylindrical holes in often sizable rocks. The stone "pestles" found are short pebbles, shaped and used, which do not seem particularly suited for use in these mortars, but one was seen in current use with a small stone mortar at a farm-house at Mochon Point on Rota. Bell-shaped limestone pounders of Caroline Islands type are also recorded from the Marianas.

The tripodal stone metate was introduced from Mexico in the seventeenth or eighteenth century, and is still to be seen in outlying villages on Guam. Metates cast of concrete by the Japanese for sale to the natives are still in use in the other islands and are found also on Guam.

Other stone objects found include well-made adzes and chisels with polished edges (similar to Melanesian rather than Polynesian types cf. Clifford, 1951), crudely chipped flint knives or scrapers, hammers and pounders, net sinkers and slingstones. Bone implements of the pre-1700 Chamorros included barbed spearheads (large: of human tibiae, as in Melanesia more recently), and "needles" or awls (for making basketry or matting, I presume). Perforated tortoise-shell disks are said to have served as trophies and currency. Many things were made from shell — fishhooks and gorges, adzes and scrapers, spoons, knives and awls, perforated shell disks (currency); shell rings are also found which are believed to have been imported from the Caroline Islands. Spears and slings were the only weapons; bows and arrows were unknown in the pre-Spanish Marianas as in the rest of Micronesia and Polynesia generally.

Social organization need not be gone into here: matrilineal clans, monogamy, a well-developed caste system of three are distinct classes, ancestor cult and shamans, are reported (Garcia's life or Sanvitores, 1683, translation in the Guam Recorder, 1936 1939).

Of the pre-Spanish (and early historic) culture the following material remains have survived to be found archeologically:

the latte stones at former village sites
stone mortars, pestles and pounders, polished stone adzes and chisels, rough hammers and chipped scrapers
pottery, almost entirely in sherds
bone awls or needles
bone spearheads (of human tibiae)
stone net-sinkers
shell fishhooks and gorges
shell adzes, scrapers, spoons
perforated shell disks; rings

For pertinent references, in addition to those already cited in this section, and general works on Oceania such as Douglas Oliver, The Pacific Islands (Cambridge,1951) or Felix M. Keesing, Native Peoples of the Pacific World (New York, MacMillan,1946), see C. R. H. Taylor, A Pacific Bibliography, printed matter relating to the native peoples of Polynesia, Melanesia, and Micronesia, Memoir No. 24 of the Polynesian Society, Wellington, New Zealand, 1951.


1. Among the few published papers on skeletal material from the Marianas, an interesting one is F. Wood Jones, "On two mandibles from Guam, "Australian Journal of Dentistry, July 1, 1931, which describes the jaws as very massive comparable to New Caledonians, with prominent square chins. There is also his unpublished "Skulls from Guam" (ms. at the Bishop Museum), I have not read it, but a few of the series of 88 crania were seen on January 14, 1952, at the Bishop Museum, with Dr. Charles Snow of the University of Kentucky, who was then working on Hawaiian skeletal material there, — The skulls are very large and muscular, and the lower jaws are enormous. The comparison to the Heidelberg jaw made by R. W Leigh (op cit. infra) is, as Wood Jones (1931) points out, purely metrical and not morphological — and meaningless, actually — but the correspondence in measurements is quite striking.

Skeletal material from the Marianas of Dr. Spoehr's,examined cursorily at the Chicago Museum of Natural History on May 5, 1952, includes a few strongly-built men but obviously none of gigantic stature, and a few small skulls of distinctly Melanesian effect.

Wood Jones' unpublished description of Guam skulls is partly summarized in a recent paper by Edward E. Hunt, Jr., "A view of somatology and serology in Micronesia," American Journal of Physical Anthropology, 8:157-184, No. 2, June 1950. Features mentioned include "an oval or ovoid vault, broad zygomatic arches [cheek-bones] * * * and well-defined temporal crests. Although aglabellar eminence [forehead prominent above the nose] is often seen in the males, the brow-ridges are usually not strongly developed. The palate is broad and the teeth large. The mandible is generally massive."(Hunt, 1950, 161-162.)

Another specific reference on Chamorro skulls is: Otto Schlaginhaufen, "Ueber eine Schaedelserie von den Marianen," Jahrbuch 1905 der St Galllschen Naturwissenschaftlichen Gesellschaft, 454-509, 1905 (not seen — cited by Hunt). Schlaginhaufen measured 23 old Chamorro skulls from Saipan; the series showed considerable range in cranial index, forward and lateral projection of the malars (cheek-bones), slight alveolar prognathism (protrusion of the jaws), and low and rather flat noses (Hunt, 1950, 161).

Among the Japanese papers on Micronesian physical anthropology cited by Mr. Hunt, only one is concerned with skeletal material from the Marianas (and from Palau and Truk also — Seiji Arai, "On the cranium and extremity bones of Micronesians," Kagaku Nanyo 4:1-14, 1941). Hunt says that Arai found that the Chamorro and Palau skulls were larger and broader than those from Truk, and noted the strong temporal and neck muscle attachments, narrow foreheads and nasion depression (the root of the nose low, under a prominent glabella), and massive lower jaws. Finally, there is also a specialized study published on teeth from the Chamorro skulls: R. W, Leigh, "Dental Morphology and pathology of prehistoric Guam," Memoirs of the B. P. Bishop Museum 11:257-273, Honolulu, 1929. Very little caries (dental decay, resulting in cavities) was observed, but teeth had been lost frequently over the age of about 35.

The skeletal material from Guam in the Bishop Museum is currently being studied by Professor Riesenberg of the Anthropology Department of the University of Hawaii, and a full report may be anticipated.

The present-day people of the Marianas are quite different, with only a few gigantic individuals of the old Chamorro type surviving in southern Guam. The native population was reduced from perhaps 50,000 to about 4,000 during the last quarter of the seventeenth century, and a considerable number of soldiers and settlers from the Philippine Islands were brought in by the Spaniards in the eighteenth and nineteenth centuries. The modern Guamanians and Saipanese are a mixture of Chamorro (and Carolinian, surely), Spanish, Mexican Indian, and Filipino, with the last the predominant element. Many of them would not look out of place in the Southwest or northern Mexico.

Hunt's paper (loc. cit.; 1950) is largely concerned with living Micronesians. He first discusses the anthropometry and metric types, on the basis mainly of K. Hasebe, "The natives of the South Sea Archipelago [Micronesia]," Jinruigaku Senshigaku Koza [Anthropology and Archeology Lectures] 1:1-35, Tokyo, 1936 (in Japanese; a manuscript English-translation by SCAP available to Hunt, not seen by me), and summarizes Hasebe's theory of Micronesian racial history, involving several migrations. Hunt then takes up blood-group distributions (not used by Hasebe at all), other recent research and current theories on Micronesian anthropology.

The major points relating to the Marianas in Hunt's survey are weakened, I believe, by underrating the importance of post-1700 Spanish, Mexican, and Filipino elements: "The craniological evidence suggests that the aboriginal Chamorros were related to the other high islanders in western Micronesia [p. 162;] In the living, the western high islanders, including the Saipan Chamorros [who are of Guamanian type and derivation are more Mongoloid than (the others) today. Indeed, the Chamorros now are metrically the most Mongoloid group in all Micronesia [p. 163 — this may be misleading or meaningless except, perhaps, to anthropologists — although many skull measurements are close to those of Orientals, the people do not resemble Chinese, Japanese, or Mongol types, with the exception of occasional individuals; but they are very like Filipinos and some Mexican types, and hence are "more Mongoloid" than the distinctly non-Mongoloid people of the Carolines and Marshalls and Polynesia]. How much of this specialization comes from recent admixture with Filipinos, Mexicans, and others, however, is not yet certain. [p.164 — obviously much of it is, from historical data and eighteenth-century population figures] * * * The Chamorros of the Marianas are metrically the most Mongoloid group, although they do not have the high percentage of gene q (the "B" agglutinogen) which usually characterizes Mongoloids nearer Asia [p. 180 — actually, as I understand it, blood-group B is distinctively characteristic of Mongoloids and others in Asia, and a very low percentage or entire absence of B appears in the Philippines as well as in Polynesia generally among American Indians other than Eskimo]."

Micronesia and Melanesia are regions of short stature, according to H. L. Shapiro (The Physical Characteristics of the Ontong Javanese, APAMNH 33-3, 1933; and The Anthropometry of Pukapuka, APAMNH 38-2, 1942). The average statures of groups of males in Melanesia ran up only to 171 cm. (5'7-1/4", in Fiji, where there is Polynesian admixture), and the Micronesian means of stature of male series from the Carolines, Marshalls, and Gilberts, "fall generally between 160 and 165 cms." (or, only 5'3" to 5'5" — Shapiro, Pukapuka, p. 167).In contrast, Polynesian male group means for the most part range between 169 and 172 cm., or 5'6-1/2" to 5'8", with a few higher ones (a Tonga series 173 cm., Easter Island between 173 and 174, and Rarotonga 174 to 175) and one markedly smaller (Pukapuka 166 cm., or 5'5-1/4').

Unfortunately, I have no figures on stature in the Marianas, for either modern Guamanians or early Chamorros. The present-day people are, with rare exceptions, of small stature; at least some of the original people were very tall like Polynesians. There are, however, a few early references to small very dark people who formed the lowest caste.

Of several, possible theories to explain the shorter darker folk, the obvious reasonable and likely assumption is that a Melanesoid or typical Micronesian population (the few Marshallese an Palauans I've seen were small, or of moderate stature, and very dark, with rather Melanesian or Australoid features) was dominated by a group of tall and heavy brown-skinned Polynesians, the Chamorros proper. A separate later arrival of this ruling class is an equally obvious theory, for which there is no known evidence whatever; it is possible that the short dark slaves accompanied the Chamorros proper in a single immigration, if it was from a source to the south or southwest, and if it was sufficiently late that the Melanesians (?) had not been absorbed, had not amalgamated with the Polynesians, by the sixteenth century; it is also highly possible that they derived from occasional Carolinian voyages to Guam, such as are historically documented for 1721, 1788, and 1804 onward. This would fit with the occurrence archeologically of artifacts of Caroline island types (bell-shaped stone pounders, Conus shell rings — Thompson, 1932, pages 41 and 57), which could also, of course, derive from the eighteenth-century or even the nineteenth-century Carolinian visits and small-scale immigrations.

In any case, at least a major and typical group of the original Chamorros evidently were of Polynesian rather than Oceanic Negroid or Oriental Mongoloid type, tall and heavy, with light or medium brown skins, black or very dark brown hair and eyes, no doubt the soft, fine, wavy to curly hair of Polynesians rather than either the straight, heavy, Mongoloid type of black hair or the sharply curved Negroid "frizz1y" variety, and with the "Caucasoid" (actually, Ainoid-Australoid) features of many Oceanic peoples.

A description by an Englishman in 1686 is more helpful than most early travelers' comments: "The natives of this island [Guahan] are strong-bodied, large-limbed. They are copper-coloured, like other Indians; their hair is black and long, their eyes meanly proportioned; they have pretty high noses; their lips are pretty full, and their teeth indifferent white. They are long-visaged, and stern of countenance, yet we found them to be affable and courteous." (William Dampier, A New Voyage Around the World, Vol. I, Ch. 10.)

The first description of Palauans is also interesting and pertinent: "The natives of these islands are a stout, well-made people, rather above the middling stature; their complexions are of a far deeper colour than what is understood by the Indian copper, but not black. * * * Their hair is long and flowing, rather disposed to curl" (George Keate, An Account of the Pelew Islands * * * Capn Henry Wilson, * * *, London, Second Edition, 1788, p. 318).

ADDENDUM: A newly-published paper bearing on the subject is: R. T. Symmons, J. S. Graydon, N. M. Semple, Joseph B. Birdsell, John D. Milbourne, and J. H. Lee, "A collaborative genetical survey in Marshall Islanders," AJPA 10:3l-54, No. 1-March 1952. In this collaborative study, Dr. Birdsell has summarized the several theories on the racial affiliations of Micronesians as follows:

  1. W. W. Howells — allied to Polynesian with a strong Indonesian influence (a brief comment in his Mankind So Far, New York, 1944; E. A. Hooton, in Up from the Ape, New York, 1946, is more discreet and does not mention Micronesians specifically);

  2. Hasebe (Hunt, 1950) — combination of negroids, a "small generalized" type, Polynesians and Indonesian Mongoloids;

  3. Birdsell — negritoid, "Caucasoid" (Ainoid), and Mongoloid basic components;

  4. J. Avias ("Les groupes sanguins des Neo-Caledoniens et des Oceaniens en general du point de vue de l'anthropologie raciale," L'Anthropologie 53:209-239, 434-477, 1949) — mixture of Melanesian, Indonesian, Polynesian, and recent Asiatic Mongoloid elements.

The results of blood-group testing, as given in this article, do not support the suggestions of Indonesian and Mongoloid and Polynesian elements or admixture, and show little difference between Marshallese and Papuans (Melanesians).

<<< Previous <<< Contents >>> Next >>>

Last Updated: 14-Feb-2004