Sea People

by Christina Thompson

  Of course, quite a lot was already known about the peoples of Polynesia. Anecdotal information about everything from Tahitian dietary habits to the monolithic sculptures of Easter Island could be culled from the journals of countless explorers, missionaries, memoirists, and travel writers. But it was all rather scattered and fragmentary and, above all, unscientific. What was wanted was detailed, systematic studies that assembled large bodies of information in ways that made them easy to classify and compare. The plan of the Bayard Dominick Expedition (actually a series of expeditions) was to describe a number of Polynesian cultures as completely as possible. Four teams of scientists, including ethnologists, archaeologists, and botanists, were sent to four different parts of the Polynesian Triangle: Tonga in the west, the Marquesas in the east, the Australs in the south, and Hawai‘i in the north. Most of the participants were in their late twenties or early thirties; some were still graduate students; several were accompanied by their wives, who were listed officially as “volunteers.” Each of the groups was to stay put for at least nine months, during which time they were to gather as much information as possible—for if the Bayard Dominick Expedition was anything, it was an exercise in data collection.

  In practice, the brief of the expedition was so broad that there was room for members to investigate whatever they wanted. The anthropologist Edward W. Gifford made an extensive study of Tongan social structure and customs relating to law, property, religion, and war. Ralph Linton, an archaeologist with the Marquesan team, examined stone walls, terraces, platforms, sculptures, petroglyphs, and fortifications. Robert Aitken, in the Australs, studied mythology, while the botanist Forest B. H. Brown and his wife undertook a survey of the flora of southeastern Polynesia, with particular attention to local botanical knowledge and nomenclature.

  One member of the Bayard Dominick Expedition was the ethnologist Edward S. C. Handy, who, with his wife, Willowdean, spent nine months in the Marquesas. The two of them made a formidable pair, he focusing much of his attention on Marquesan myths and religious practices while she made detailed studies of string figures and tattooing. Although the practice of tattooing had been outlawed by the French government in the late nineteenth century, Willowdean managed to find more than a hundred Marquesans with significant tattoos. She persuaded them to let her document these extraordinary designs, meticulously copying them onto paper when she discovered that the blue-black markings did not show up well in photographs. Years later, she recalled what it had been like to drop down into the middle of a Polynesian society in 1920. Much of her energy, she wrote, had been spent trying to figure out the social behaviors—what was considered polite, what was offensive, what was the true meaning of the things people said and did. But part of the learning curve had to do with her own shifting perspective: she had come to the island, she wrote, thinking of its inhabitants “as repositories of information” and departed nine months later thinking of them as friends.

  Still, the purpose of the expedition was to collect information, and the Handys dutifully went about this task, inquiring about dances and dress styles, witchcraft, fish poisons, boxing, bird catching, ideas about time—anything and everything they could think of. They were especially interested in Marquesan legends, which seemed to Willowdean unlike anything she had ever encountered, with a louche, casual violence and a sexuality so explicit that some of the stories were “unpublishable” in English. Marquesan myths had gone underground in the 1920s in much the same way as tattoos, and many of the stories the Handys collected during this period were recounted by a man known locally as “a great liar,” meaning, wrote Willowdean, that “he told the stories of the old-time people which the missionaries called ‘lies.’”

  Each of the Bayard Dominick teams was expected to complete a “physical survey” of the islanders and was provided with a special set of instruments for the purpose: calipers for measuring faces and heads, standing rods for determining height, pigmentation charts for classifying skin color, and special “anthropometrical cards” on which to write it all down. The data, which included numerical codes for attributes like hair form and skin tone as well as a standard set of body measurements, were to be gathered and sent, along with hair samples and photographs, to the expedition’s “somatologist,” Louis R. Sullivan of the American Museum of Natural History, who was tasked with compiling and publishing the results.

  The Handys quickly realized that they would need to measure every full-blooded Marquesan who came their way if they were to meet their quota of three hundred “samples.” Edward busied himself with calipers and standing rod while Willowdean wrote down the numbers on a chart, which, as she put it, “covered every feature of the human body from the crown of the head to the soles of the feet.” Then she would snip off a bit of hair (she was surprised that the Marquesans allowed her to do this, since hair was known to be used by sorcerers “as bait to entrap the spirit of a man”), paste it onto a card, and try to match the subject’s skin color to a scale of pigmentation shades. This last, as it turned out, was one of the most difficult parts of the job.

  Several different systems existed for assigning skin tones a numerical value. They included Broca’s Couleurs de la Peau et du Système Pileux (Colors of the Skin and Body Hair), published in Paris in 1879; a color panel designed by the German anatomist and physiologist Gustav Fritsch, consisting of forty-eight shades painted in oil on special paper in a handy pocket-size case; and Felix von Luschan’s Hautfarbentafel, or skin-color panel, a seven-by-three-inch brass tray containing thirty-six glass mosaic pieces in shades from ivory to nearly black, which also came in a smaller, double-sided version fitted with a sliding brass sleeve. Von Luschan’s Hautfarbentafel was considered the gold standard, but it was not available to the Bayard Dominick teams, who used Fritsch’s kit instead.

  Willowdean was often flummoxed by her assignment. At first, she thought it would be easy with the help of a pigmentation chart, but all too often she could not identify the right number. “Everyone complains about being unable to match Polynesian skin colors with the standard chart,” her husband told her. “Put down the number of the nearest match.” Sometimes she would put down two numbers with a plus sign between them. “Once,” she wrote, “I was tempted to write a fraction.” Held up against the cheek of a living person or against the inside of the upper arm—the two preferred locations for measuring “exposed” and “unexposed” skin—the flat, painted colors of the little paper blocks were woefully inadequate. Most of the Marquesans, Willowdean wrote, “had skins that were glowing, golden, light, yellow, brown—how describe them? Copper-colored under sunlight?” She wondered how an artist would handle the situation: “Would a painter mix chrome yellow with sepia, perhaps a bit of rose madder?” The adjectives that came to mind were all poetic, “but we were dealing with scientific data. I had to put down the approximate number.”

  SOMATOLOGICAL STUDIES—THAT IS, studies of the human body such as those undertaken in Polynesia by the anthropologists of the Bayard Dominick Expedition—were made of many different classes of people in the late nineteenth and early twentieth centuries. Typical subjects included women, children, athletes, “mental defectives,” ethnic subgroups like Scandinavians and Eskimos, twins, Neanderthals, and the criminally insane. There were studies of distinctive physical features—palms, soles, teeth, ears—and a great deal of interest in skulls, crania, and the brain weights of various types of people, including artists, scholars, scientists, and “distinguished educated men.” The ostensive purpose of all this activity was classification, the establishment of the range of variation, and an exploration of the mechanisms by which variation occurred.

  Some somatological studies were purely descriptive—studies of the angle of the elbow, for instance, or the comparative growth rates of girls and boys. But many were driven by sociological questions: attempts, for example, to establish links between physical attributes (like stature or head shape) and mental states (like insanity) or social conditions (like poverty or being fi
rstborn) or cultural customs (like nomadism). The idea that complex questions about the human condition could be answered by measuring people’s bodies was not new. Johann Blumenbach’s classic eighteenth-century division of humanity into five distinct types (Caucasian, Mongolian, Malayan, Ethiopian, and American) had been based on craniometric data—taken, apparently, from his own extensive collection of skulls. But in the latter half of the nineteenth century, debates about evolution and heredity had ignited interest in the notion that heritable physical characteristics could be definitive when it came to human populations. What was needed to prove this was data, and many large-scale studies were undertaken (including a census of the eye and hair color of six million German schoolchildren, which gave rise to a rumor that the king of Prussia had staked and lost forty thousand fair-haired children in a card game with the sultan of Turkey).

  One of the many pernicious aspects of this research was the way it was used to classify human “races.” In biology and anthropology, race has long been abandoned as a meaningful category; indeed, although the term has been in use since the seventeenth century, it has never been precisely defined. There has never been any agreement about the number of human races, or what the definitive characteristics of a race might be: skin color, hair type, head shape, etc. Genetic research in the twentieth century has shown that there are no genes that correspond to racial types and that the range of variation within so-called races is greater than the variation across them. But scientists in the early 1920s were working with an essentialist model of race as something immutable, definitive, and grounded in biological reality.

  The framework within which the anthropologists of the Bayard Dominick Expedition operated was that there existed a certain number of “pure” human races. The minimum was generally considered to be three: Caucasoid, Mongoloid, and Negroid. But many of the world’s peoples did not fit clearly into any of these categories, and scientists frequently invented additional racial types—Malayan, Indonesian, Austronesian, Negrito—or argued that these unclassifiable people represented populations that were “racially mixed.” Polynesians (along with Native Americans, Melanesians, and Australian Aborigines) were one of the ambiguous groups, and one goal of the Bayard Dominick Expedition was to ascertain, using anthropometric data, what unique medley of existing races had “combined to make the Polynesian physical types.”

  It had long been observed—indeed, it is a cliché of early descriptions of Polynesians—that the islanders of the remote Pacific all looked remarkably alike. And yet, beyond the fact that they were tall, strongly built, dark-haired, brown-skinned, and generally handsome, there was little consensus among Europeans about how to describe them, especially when it came to the color of their skin. Nor could Europeans agree about who else in the world they most closely resembled, with some saying they looked like American Indians and others that they were more like Indonesians or Filipinos. Ideas about Polynesians’ racial makeup were equally muddy. Many scientists, influenced by the still powerful Aryan theory, considered them Caucasian. Others described them as Mongoloid (i.e., Asian), while a third group identified them as belonging to the same stock as the indigenous peoples of the Americas—though, as no one was quite clear what race they belonged to, this was hardly a helpful idea. Most, however, concluded that Polynesians were “a hybrid people,” a mixture—or layering, as one theorist imagined it—of Caucasian, Mongoloid, and Negroid elements in some combination.

  Sullivan, whose job it was to crunch the data delivered to him by the Bayard Dominick field teams and bring some clarity to the situation, had a difficult task. The data sets ran to thousands of numbers: hundreds of subjects analyzed in terms of dozens of features. To begin with, there were the body measurements: stature, shoulder height, arm length, sitting height, head length, face width, nasal height, ear width, as well as the “bigonial diameter at the angles of the mandible.” Then there were all the nonquantitative features: skin tone, hair form, eye color, tooth shape—all of which were assigned numerical values. Sullivan also calculated a whole series of ratios with exotic-sounding names like the “transverse fronto-parietal index” and the “zygomatico-mandibular index” and tried grouping subsets of the data in all different ways, noting the hair form and cephalic index of the men with the longest arms, or the stature and pigmentation of women with the shortest heads, seeking, quite hopelessly, as it turned out, to identify significant clusters or correlations.

  The results were presented in a series of studies based on anthropometric data from Samoa, Tonga, and the Marquesas and published by the Bishop Museum between 1921 and 1923. (A fourth study, of Hawaiian data, was in preparation when Sullivan died, prematurely, in 1925.) The conclusions are complex—inscrutable, even—but it is interesting to trace the evolution of Sullivan’s thinking as he worked his way through this mass of statistical information.

  Beginning with the observation that Polynesians were commonly described as having “European racial affinities” (a gesture toward their supposed Aryan roots), Sullivan first concluded that the evidence from the Bayard Dominick Expedition contradicted this. Based on his analysis of Samoan and Tongan data, he found that, on the contrary, they were “most closely allied to the Mongoloid race of mankind”; in other words, they were essentially Asian. In the case of the Samoans, Sullivan found that, to the extent that they diverged from this type, it was in the direction of the European. When it came to the Tongans, however, the variation tilted in a “Melanesian” direction, meaning that Tongans were a little more black. (It is worth noting that Tongans were considered one of the “purest” Polynesian populations. Tonga, which is the only island nation in the Pacific never to have been ruled by outsiders, had a population in 1920 of more than 23,000 Tongans and only 347 Europeans. Compare this with the 1920 census for Hawai‘i, which lists 41,000 Hawaiians, 54,000 Europeans, and more than 150,000 combined Filipino, Japanese, and Chinese.)

  Sullivan next turned his attention to the Marquesas, where the situation was complicated by the cataclysmic population collapse of the nineteenth century and the high rates of intermarriage between the Marquesans and both Europeans and Chinese. Here, perhaps under the influence of Edward Handy, Sullivan introduced an entirely new model. The Marquesans, he decided, could be divided into two types: one characterized by taller stature, longer heads, narrower noses, straighter hair, more body hair, and lighter skin color (all classic “Caucasian” markers) and the other by shorter stature, broader heads, wider noses, wavier hair, less body hair, and darker skin color (all stereotypical “non-Caucasian” features). To the first he gave the name “Polynesian”; to the second, “Indonesian.” This led Sullivan to rethink his earlier analyses of Samoa and Tonga, and, with the benefit of hindsight, he now realized that these two types—a taller/lighter, and a shorter/darker—had been present all along in all these other populations. The upshot of all this was to shift the primary racial category of the Polynesians back to Caucasian from Mongoloid, making them once again “white” and Indo-European rather than Asian, while still allowing for a “blackish” element or strain.

  WHAT IS PAINFULLY clear from Sullivan’s attempts to rationalize all this anthropometric data—so many numbers, so laboriously collected, so tirelessly grouped and regrouped—is how entirely futile it seems to be. In fact, no analysis of head shape or nose width or cephalic index was ever going to produce anything but a murky result. No tabulation of pigment would ever correlate with arm span or ear height in the way Sullivan hoped, because, even if the patterns he was looking for were there, the tools available to him in the 1920s were incapable of revealing them. All of Sullivan’s calculations had to be done by hand, and neither the statistical methodology nor the computational power that would have enabled him to extract meaningful results from this data had yet been invented. And so instead, what we see is a jumble of results reflecting not some truth about the data but a set of underlying assumptions about the people themselves.

  In his Marquesan study, Sullivan directed the re
ader to a group of photographs at the end of the volume that had been sorted into his “Polynesian” and “Indonesian” types. These photos, consisting of head shots of men and women, face-on and in profile, resemble nothing so much as mug shots or passport photos. The subjects stare at the camera with serious expressions, some look a little worried or annoyed, a few wear the tiniest of smiles. Although many appear to have dressed up for the occasion, you could never mistake these images for portraits. They are clearly specimen studies, designed to be as consistent and uniform as possible in order to facilitate comparison and classification. Sullivan’s somatological studies were published in large quarto volumes, allowing for as many as twelve, fifteen, even sixteen head shots per page, and this presentation encourages the reader to judge for himself whether what Sullivan said—that there can be “no doubt that at least two separate and radically different groups . . . are represented”—is borne out by the photos.

  And the truth is, when you look at these arrays of Polynesian faces, you really cannot see what Sullivan was talking about. All of these people look sort of alike, and yet, as one would expect, they are also all different. One has a longer nose, another a broader; one’s hair is wavy and a bit disheveled, another’s is straighter, perhaps more carefully combed. Looking at these pictures, one cannot help seeking a sense of the person behind the shot. Who were these men and women? What were they thinking? Willowdean Handy wrote that among the Marquesas there was great demand to be photographed and that each of the sitters received a print. (The Marquesans, it seems, did think of these photos as portraits.) And other anthropologists have also described these measuring and photographing sessions as popular social events, with lots of joking and banter from the assembled crowd. None of this comes through in the photos, however, and it is interesting to speculate about what else has been eliminated in the drive to achieve strict standardization. But even putting aside one’s natural human sympathy and viewing these images as they are meant to be viewed—as type specimens—it is impossible to see in them any of the classificatory systems Sullivan described.