The Man-Eating Myth Reconsidered

by C. R. Hallpike (August 2018)

Cannibals Preparing Their Victims, Francisco Goya, 1800-1808




Professor William Arens’s book The Man-Eating Myth was published by Oxford University Press in 1979, and claims that cannibalism is a racist and colonialist myth perpetuated by Westerners, including credulous anthropologists who should know better, and that there is no reliable eye-witness evidence that it ever existed as a social custom in any society (as distinct from occasional “survival cannibalism”). The book created something of a sensation when it appeared and, although we are approaching its fortieth anniversary, it is still in print, with respectable sales on Amazon and discussed at great length in Wikipedia, and so seems worth a further assessment. Besides telling post-modernist academia what it wanted to hear, it has clearly satisfied a popular need as well, about which the following extract from a review on Amazon gives us a clue:


The reason this book caused such a ruckus when it was released, is not just the fact that it made anthropologists look as disreputable as phrenologists: charlatans, shysters and hucksters practicing a crank pseudo-science. Among the highly educated, it’s fashionable to ridicule the bumpkins and yokels for being gullible enough to buy into astrology, creationism and other forms of nonsense. But as W. Arens proved with “The Man-Eating Myth”, the intelligentsia is just as easily fooled as what Mencken called “the booboisie” [1] and that in many cases, “PhD” means “piled high and deep”.


It is undoubtedly true that cannibalism is the feature of primitive society most apt to be sensationalised by the popular press in particular, and books with titles like Where Cannibals Roam, A Naturalist in Cannibal Land, The Last Cannibals, Mountains, Gold and Cannibals, or Two Years Among New Guinea Cannibals are sure to find eager readers. Anthropologists would also agree that many accounts of cannibalism are exaggerated, based on rumour, or simply false. Probably all societies have contrasting images of the wild and the tame or social: standard images of the wild are incest as opposed to respect for kinship rules, eating food raw as opposed to cooked, nudity as opposed to clothing, hairiness and long hair as opposed to smooth skin and short hair, and eating human flesh as opposed to animal flesh, so it is not surprising that accusations of cannibalism are often used to stigmatize “the other.” For example, the Konso of Ethiopia (Hallpike 2008) had a horror of cannibalism, and a very old man told me that in his youth he had been to Addis Ababa (about four hundred miles to the north) on an errand for the Imperial Government. He stayed there for some time and, on his way back, he was misdirected about the road. After walking “for a year,” he reached the land of the cannibals, the pulkoota. Their mouths, he said, “stuck out like this”—holding his fingers towards his mouth and clearly indicating an ape-like face—and they had eyes in the backs of their heads. They used to buy people and also kept prisoners captured in battle. They would cut them up into strips and hang these up to dry. They lived only on human flesh and cultivated no fields. He managed to avoid them and eventually made his way back to Konso (ibid., 379). And when I first began living among them, some of the mothers would tell their children that if they did not be quiet and go to sleep the terrible white man would come and eat them. The Konso conception of cannibalism is an excellent example of a pervasive theme of Arens’s book, that cannibalism is a stigmatization of the savage “other”.


If this were all that Arens is saying it would be accepted as a commonplace of anthropology, but he raises the commonplace to the sensational by claiming that there is no evidence that cannibalism has ever existed at all: “. . . excluding survival conditions, I have been unable to uncover adequate documentation of cannibalism as a custom in any form for any society. Rumors, suspicions, fears and accusations abound, but no satisfactory first-hand accounts.” Although we may find this extremely surprising, he nevertheless goes on to assure us that “I have marshalled the available material to support this premise, rather than manipulating the data to generate the kind of foregone conclusion which characterizes the current thinking on this topic” (Arens 1979:21-22).


Before we go any further, however, it is very striking that Arens never makes any attempt to explain why the refusal to eat human flesh must apparently be such a powerful and universal human imperative that cannibalism has never existed anywhere as an accepted social practice. He simply assumes it to be self-evident. One might be unwilling to believe, in principle, that any society could possibly have institutionalised incest between mothers and sons, or the eating of human faeces, for example. But in primitive societies (small-scale, face-to-face, non-literate, with subsistence economies) especially, meat is highly prized particularly by those dependent on agriculture because they can only eat it relatively seldom. Since people in many such societies are willing to eat stinking meat, why is it inconceivable for them to eat fresh human meat, especially of enemies killed in battle? Indeed, symbolic cannibalism is quite familiar to Christians when they take the sacraments of Christ’s Body and Blood. Arens’s unwillingness to believe in the very possibility of cannibalism as an institution appears, in fact, to be his own ethnocentric Western prejudice.


His demand for eye-witness evidence begins with the undoubted fact that “cannibalism is an observable phenomenon” but proceeds to the very dubious inference that therefore “the evidence for its existence should be derived from observation by reliable sources” (21), meaning “those trained in the craft of ethnography.” There are in fact plenty of eye-witness accounts of cannibalism: “Claims of having observed cannibalism first-hand are rampant in the travelogues of explorers, missionaries, explorers, sailors and their ilk” (35). But he dismisses all these as having “little if any credibility,” and continues


Leaving this brand of literature behind, and examining instead the production of professional anthropologists, the problems change but the situation still remains perplexing. From all corners of the globe the reports come in that a specific group of people an anthropologist has lived among were cannibals long ago, until pacification, just recently or only yesterday. The reader is engulfed by a stream of past tenses denoting varying removes in time, indicating a demise of custom some time before the researcher took up residence upon the scene (35-6).


With one trivial exception of the ritual consumption of human ashes, which he rejects, he claims that no modern anthropologist has witnessed the consumption of human flesh by the people he or she was studying. This is not really perplexing at all, of course, since colonial administrators and missionaries had suppressed cannibalism so, by the time the anthropologists came on the scene, they were too late to observe it. But not only does Arens dismiss eye-witness accounts by those not “trained in the craft of ethnography,” but is almost equally contemptuous about anthropologists generally and New Guineanists in particular, where apparently “. . . academic standards seem to function as an almost forgotten ideal, rather than as standard operating procedure. Anthropologists with well-deserved reputations based upon previous research and publication become the victims of their own sensationalism and poor scholarship” (99).


The anthropologist Klaus-Friedrich Koch, for example, supplies copious details of cannibalism among the Jale of West New Guinea (Irian Jaya), such as:


Cannibalism is an integral part of a particular kind of war. The Jale distinguish between a wim war and a soli war. Only soli warfare ideally features anthropophagic revenge. While a wim war always ends within a few years and may last only for a day or two, a soli war usually endures for a much longer time and may extend over the period of a generation . . . Wim warfare occurs between two or more wards of the same village, between two segments of the same ward living temporarily at different localities such as garden hamlets, or between two or more villages in the same district or adjacent districts. Soli wars, on the other hand, are usually waged between two villages separated by a wide river or by a mountain ridge, a geographic condition that puts them in different districts or regions. Informants repeatedly stated the maxim that ‘people whose face is known should not be eaten’. In practice immunity from anthropophagic vengeance derives from the nature and relative frequency of affinal links between two villages (Koch 1974: 79-80).


Arens, however, simply dismisses all Koch’s research as the result of missionary propaganda, since he cannot claim to have witnessed cannibalism himself (Aren 1979: 98) but fails to ask himself the next and perfectly obvious question “Where, then, did Koch get all this stuff about cannibalism—did he just make it all up?” Obviously he didn’t, and got it from his informants as he makes abundantly clear in his book but it would be inconvenient for Arens to admit this since, as we have noted, it is one of the themes of his book that accounts of cannibalism are inherently hostile and derogatory lies told about other peoples, and not about one’s own.


Why, however, would the Jale say they had been cannibals if they hadn’t, and why did the Tauade happily admit to me that they had been cannibals too, referring to enemy groups with a laugh as “our meat, like pork.” In my book Bloodshed and Vengeance in the Papuan Mountains (1977) I give the following account of a cannibalistic event told to me by my best informant, Amo Lume:


While the initiation ceremony was in progress the Gane men made an attack. The Goilala seized their weapons and chased the Gane. There was a big battle. Aima Kamo speared Kog Kanumia Konoina, and Aima Kovio also speared him, and Koupa Teva axed him, as did Orou Keruvu, and Mo Kimani, chief of Watagoipa. Everyone came and chopped him to pieces. The Tawuni and Kataipa, valavala [allies] of the Goilala, were invited to take the bits home to eat. Kolalo Kioketairi (who had a twisted lip because he had cut his mouth while removing human flesh from a bone) cut off Kog Kanumia’s head and took it to Dimanibi singing a song. [Then the storyteller retraces his steps to give some further details.]


 . . . the Tawuni and Kataipa took away their pork [given by the Goilala to celebrate the victory] with Kog Kanumia of Gane’s body. They dismembered Kog at the Kovelaiam bridge over the Kataipa river, and made a big oven [an earth oven with hot stones], in which they cooked the pork and Kog Kanumia at the same time. Kolalo tied a vine to Kog Kanumia’s head and held it over the fire to singe off the hair [pigs similarly have their hair singed off before cooking], then cooked it in the oven. When it was taken out, he skinned the face and feasted on the white flesh beneath. After this the Tawuni and Kataipa went back to their places. (Hallpike 1977:213).


The remark about “the white flesh” beneath the dark skin of the face is interesting, because in fact even dark-skinned human flesh, when cooked, does indeed turn white, like pork and chicken, a realistic detail which gives additional credibility to the story. Arens dismisses my account of Tauade cannibalism, carefully ignoring the episode just described, but again fails to answer the basic question of why on earth the Jale, the Tauade, and many other peoples of Papua New Guinea would claim to have been cannibals themselves in the past if this were not true.


In the same way as these accounts of cannibalism from native informants, anthropologists have had to rely on the people’s recollection of other aspects of their life and culture that were suppressed or had died out, like warfare or exposing corpses to rot, or initiation ceremonies, but should it be assumed that native informants were lying or mistaken about all these as well? It seems a curiously disrespectful attitude to indigenous peoples to dismiss all their recollections of their own past as unreliable. The Tauade used to be one of the most violent societies on record, and my informants gave me copious accounts of all manner of warfare and mayhem, which were supported by government records but, during the two years I lived with them, I never witnessed a single homicide apart from an accident or even a physical assault, let alone a battle, yet these are all highly observable phenomena nonetheless.


So the reason that so many anthropologists’ accounts of societies in Papua New Guinea mention cannibalism is not because they had ‘become the victims of their own sensationalism and poor scholarship’, but because their informants told them a great deal about it. By contrast, a survey of the historic literature and modern ethnography of the Cushitic-speaking peoples of the Horn of Africa, which include the Konso, reveals virtually nothing on the subject of cannibalism, except one or two vague references in the earlier literature. This is not because anthropologists working in this area were more objective than those working in New Guinea, but simply because there was probably little or no cannibalism in the Horn of Africa.


At this point it is time to revert to Arens’s “basket of deplorables”—all those travellers, missionaries, and old sea-faring men he so despises. According to Arens, “The legion of reports by non-specialists were found to range from highly suspect to entirely groundless when viewed from the perspective of objective scholarship and common sense” (181), whereas they actually provide some of the best material on cannibalism. Arens’s complete denial of cannibalism puts him in the same logical position as someone who insists that all swans must be white and that accounts of black swans are absurd myths only believed by the ignorant and credulous. It simply takes one example of a black swan for the whole theory to start unravelling—if one, why not others? My black swan for Arens is the following eye-witness account of Maori cannibalism by Captain Cook. It can most easily be found in J.C. Beaglehole’s standard and readily available The Life of Captain James Cook (1974) by looking in the index under “Cook, James, reflections on cannibalism,” which took me all of five minutes to unearth. Beaglehole takes the account from Cook’s Journal for 23 November 1773 in Queen Charlotte Sound, New Zealand, and it reads as follows [2]:


There had been rumours of a war expedition to Admiralty Bay, lately picked human bones had been found, when on 23 November, with Cook anxious to get to sea but prevented by the wind, some of the officers went on shore to amuse themselves and were confronted by the remainders of a cannibal feast. The broken head and the bowels of the victim were lying on the ground, his heart was stuck on a forked stick fixed to the head of a canoe. Pickersgill gave two nails for the head and took it on board, to the interest of a number of New Zealanders on board who had not participated in the banquet. Would one of them like a piece? asked Clerke, ‘to which he very chearfully gave his assent’; Clerke cut a piece and broiled it in the galley, and the man devoured it ravenously. At that moment Cook, who had been absent, came on board with Wales, Forster, and the young islander Odiddy [not a Maori], to find the quarter-deck crowded and excitement general. Revolted as he was, the spirit of science triumphed, he must be able to bear witness from his own eyes to a fact that many people had doubted on the first voyage reports; Clerke broiled another piece, it was similarly consumed before the whole ship’s company; some were sick; Odiddy, first motionless with horror, burst into tears and abused Clerke as well as the New Zealanders, up till then his friends; Wales and Cook thought it over. (Beaglehole 1974: 358-59).


Captain Cook is renowned as one of the most meticulous and objective of observers, and it did not take someone “trained in the craft of ethnography” to describe this particular incident. But if one finds a black swan it could hardly be the only one and statistically one would expect that a number of others also existed. The most effective method of proving that something like cannibalism does not exist, would be to find cases where the evidence for it seemed to be the strongest, and then try to demonstrate that in fact this so-called evidence is fabricated or otherwise too weak to prove the case. If the strongest cases fail to demonstrate the existence of cannibalism, then it is a reasonable inference that weaker cases are likely to fail as well, even if we cannot examine all of them.


Anthropologists, among many others, have long considered that before colonial rule the Maori of New Zealand, many New Guineans and the Fijians were cannibals, which is why I naturally went first to the records of Captain Cook. Arens, however, in “marshalling the available material,” does not mention the Maori at all—about whom there is clear evidence of cannibalism from many sources (see Jennings 2004, for example) and, in the index of his book, the Fijians rate only one mention, p.176. Turning to this, and expecting at least some discussion of their celebrated cannibalism, one finds only a reference to a Hawaiian gift shop: “Here they can purchase ‘Authentic Cannibal Forks’ made in Fiji which, the package instructs the buyer, were originally used by the chiefs, since it was tapu for such food to touch their lips. It adds that missionaries stopped the practice, and suggests instead that the owner can now use these instruments as ‘pickle forks’” (Arens, p.176). This is all the evidence that Arens can marshal on the topic of Fiji, one of the most intensively studied example of institutionalised cannibalism in the ethnographic literature.


There are many eye-witness account of Fijian cannibalism from the nineteenth century, of which one of the best known is that of William Endicott (1923) based on his experiences in March 1831, as third mate of the Glide. [3] He describes going on shore after hearing that the nearby village are celebrating the arrival of three enemy corpses, killed in a recent battle, and which had been brought back to be eaten (bakola). One of the bodies was given to a neighbouring village but the other two were prepared for the oven:


The heads of both savages being now taken off, they next cut off the right hand and the left foot, right elbow and left knee, and so in like manner until all the limbs separated from the body (see Sahlins 1983: 81-2 for confirmation of this ritual practice) . . . [After a special piece was cut from the chest for the King] . . . The entrails and vitals were then taken out and cleansed for cooking. But I shall not here particularise. The scene is too revolting. The flesh was then cut through the ribs to the spine of the back which was broken, thus the body was separated into two pieces. This was truly a sickening sight. I saw after they had cut through the ribs of the stoutest man, a savage jump upon the back, on end of which rested upon the ground, and the other was held in the hands and rested upon the knees of another savage, three times before he succeeded in breaking it. This ended the dissection of the bodies. (Endicott 1923: 62) [A fire-pit had been dug about 6 feet in diameter and one and a half deep, and lined with stones, and a large fire made in it, into which small stones were placed.]


 . . . as the bodies are cut to pieces they are thrown upon the fire, which after being thoroughly singed are scraped while hot by the savages, who sit around the fire for this purpose. The skin by this process is made perfectly white, this being the manner in which they dress their hogs, and other animal food.’ (ibid, 63).


The head of the savage which was last taken off, was thrown towards the fire, and being thrown some distance it rolled a few feet from the men who were employed around it; when it was stolen by one of the savages who carried it behind the tree where I was sitting. He took the head in his lap and after combing away the hair from the top of it with his fingers picked out the pieces of the scull which was broken by the war club and commenced eating the brains. This was too much for me. I moved my position, the thief was discovered and was as soon compelled to give up his booty, it being considered by the others he had got by far too great a share. The process of cleansing and preparing the flesh occupied about two hours. There was no part of these bodies which I did not see cleansed and put in the oven.


The stones which had been placed upon the fire were now removed, the oven cleaned out, the flesh carefully and very neatly wrapped in fresh plantain leaves and placed in it. The hot stones were also wrapped in leaves and placed among the flesh, and after it was all deposited in the oven, it was covered up two or three inches with the same kind of leaves, and the whole covered up with earth of sufficient depth to retain the heat (ibid., 63-4).


This construction of the earth oven was exactly the same as that which I observed among the Tauade. It was not due to be opened until after midnight, so Endicott went off and did not return to the house where the feast was until shortly before dawn, when he found that the feast had been going on for some time. But he was not too late, and was offered a piece of meat: “It was accordingly brought carefully wrapped in a plantain leaf as it had been placed in the oven. I unwrapped it and found it to be a part of a foot taken off at the ankle and at the joints of the toes. I made an excuse for not eating it, by saying that it had been kept too long after it was killed, before it was cooked, it being about thirty-six hours” (ibid., 66-7). (Other seamen from the Glide, who also went ashore, independently confirmed the basics of Endicott’s account, see Sahlins 2003:5).


Not everyone believed this and other accounts, and Sahlins comments:


Faced by a similar incredulity, another British captain, Erskine of HMS Havannah, was compelled to preface his discussion of Fijian cannibalism by lengthy quotations from eyewitness reports of earlier European visitors. These include accounts from the voyage of the Astrolabe (1838), the US Exploring Expedition (1840), and from the missionary-ethnographer John Hunt (1840). Erskine also prints in full the narrative of John Jackson, a seaman resident in Fiji from 1840 to 1842, which contains three detailed descriptions of cannibal feasts (pp. 411-477). (Sahlins 1979)


There are many other eye-witness accounts of Fijian cannibalism from the nineteenth century, of which Sahlins mentions, in particular:


(1) Mary Wallis, the wife of a bêche-de-mer trader, was in Fiji for about 46 months between 1844 and 1851. Her diaries (1850; 1994) record some 32 cannibal events—I may be off by one or two—in 21 different locations, many involving multiple bakola. There are also five or six more general discussions of cannibal practice; (2) Rev. Thomas Williams’ published journal (1931) reports 28 cannibal incidents (including some in editorial notes, mainly from Williams’ other writings) at 17 locations, and also includes five general discussions (cf, Williams and Calvert 1859); (3) in Rev. Joseph Waterhouse’s book on Bau (1866) there are 24 instances at ten or more places, plus ten general discussions. (Sahlins 2003:5).


Sahlins gives a general description of how cannibalistic symbolism permeated the whole Fijian way of life. It was expressed in


. . . the specific drumbeats announcing the taking of bakola [cannibal victims]; the pennants flying from the masts of victorious canoes signifying bakola on board; the ovens reserved for cannibal feasts; the special stones near the temple on which bakola were carved up; the sacred trees on which their genitals were hung; the (natural) bamboo splints used to carve human flesh and the elaborately fashioned forks used to eat it; the distinctive dances, songs and unrestrained joy with which young women, dressed in finery, greeted the return of successful warriors; the sexual orgies while the bodies were cooking; the ritual consecration of warriors who had killed and the enshrinement of their war clubs in the temples; the miserable afterlife of unsuccessful warriors, pounding a pile of shit through all eternity; the gourmet debates about body parts; the taboos on human flesh for certain persons; the cures effected by pressing cooked bakola flesh to the lips of afflicted children; the sail needles made from the bones of notable bakola and the poetry from their fate. (Sahlins 2003:4, and see also Sahlins 1983: 72-93)


Not all cannibalism, by any means, was so bound up in the culture’s religious and social life and could be quite perfunctory. Mr. William Mariner was a young captain’s clerk who was captured by the Tongans in 1806 when they seized his ship and killed most of the crew. He became a favourite of King Finow, learnt the language, and was a close and very intelligent observer of Tongan life until he managed to escape in 1810. On his return to London he was befriended by a physician, Dr. John Martin, who published an account of his experiences. During one of the many wars in which Mariner was involved, he made the following observation on cannibalism:


The following day, some of the younger chiefs, who had contracted the Fiji habits [my emphasis] proposed to kill the prisoners, lest they should make their escape, and then to roast and eat them. The proposal was readily agreed to, by some, because they liked this sort of diet, and by others because they wanted to try it, thinking it a manly and warlike habit. There was also another motive, viz. A great scarcity of provisions; for some canoes which had been sent to the Hapai islands for a supply were unaccountably detained, and the garrison was already threatened with distress. Some of the prisoners were soon despatched; their flesh was cut up into small portions, washed with sea-water, wrapped up in plantain leaves, and roasted under hot stones; two or three were embowelled, and baked whole the same as a pig (Martin 1827(I): 107-8).


Mariner notes that “When Captain Cook visited these islands, cannibalism was scarcely thought of amongst them, but the Fiji people soon taught them this, as well as the art of war” (ibid., 108-9).


Mariner also witnessed a second instance of cannibalism. Sixty men had been killed in a siege of fortress by King Finow and, after they had been dedicated to various local gods, the nine or ten bodies belonging to the enemy


. . . were conveyed to the waterside, and there disposed of in different ways. Two or three were hung up on a tree; a couple were burnt; three were cut open from motives of curiosity, to see whether their insides were sound and entire [the liver of those guilty of sacrilege was supposed to become diseased], and to practise surgical operations upon, hereafter to be described; and lastly, two or three were cut up to be cooked and eaten, of which about forty men partook. This was the second instance of cannibalism that Mr Mariner had witnessed; but the natives of these islands are not to be called cannibals on this account. So far from its being a general practice, when these men returned to Neafoo after their inhuman repast, most persons who knew it, particularly women, avoided them, saying Iá-whé moe ky-tangata, “Away! You are a man-eater” (ibid., 172-73).


Despite the initial circumstances of his capture, Mariner established very friendly relations with the Tongans, whom he clearly liked, and was an intelligent, well-qualified and fair-minded observer. Modern anthropologists are quite justified in accepting his evidence, particularly as it is supported by many other observers of the period.


Another good test of Arens’s scholarship is his analysis of accounts of cannibalism in South America, in the course of which he gives close attention to a book published in 1557 by Hans Staden, a sixteenth-century German sailor and therefore a prime target for Arens’s ridicule.


. . . Hans Staden [was] an extraordinary fellow who visited the South American coast in the mid-sixteenth century as a common seaman on a Portuguese trading ship. Through a series of misfortunes, including shipwreck, he was soon captured by the Tupinamba Indians. As a result of his ill luck, the Tupinamba have come down to us today as man-eaters par excellence (22).


Aren’s most serious charge against Staden is that he had little or no command of the Tupi language which, if true, would completely discredit his account of them:


There are also the matters of language and ability to recollect to be considered. In one instance, the narrator ruefully mentions being unable to communicate his plight to a Frenchman who visited his captors’ settlement. Apparently he had no facility in the language of his fellow European. However, Staden is able to provide the details of numerous conversations among the Indians themselves, even though he was with them for a relatively limited period. He is particularly adept at recounting verbatim the Indian dialogue on the very first day of his captivity, as they discussed among themselves how, when, and where they would eat Staden. Obviously, he could not have understood the language at the time, and was reconstructing the scene as he imagined it nine years before. The later dialogues in the text must also have been a reconstruction, since there is no indication he kept notes, even if he could write. In one scene, which stands as a testimony to Staden’s memory and piety, he repeats the psalm “Out of the deep have I cried unto thee.” The Indians respond: “See how he cries; now he is sorrowful indeed” (67). One would have to assume that the Indians also had a flair for languages in order to understand and respond to Staden’s German so quickly. In summary, there was great opportunity for a certain degree of embellishment by the author, as well as by his colleagues in the eventual publishing venture (25-6).


Donald Forsyth, a leading authority on Brazilian ethnohistory, comments:


Arens’s implication (1979:25) that, because Staden couldn’t speak the language of his “fellow European,” he couldn’t speak Tupi either, makes about as much sense as arguing that because an individual has no facility in Russian, he couldn’t possibly have any in Portuguese either (Forsyth 1985:21).


It is actually obvious from Staden’s own account that he understood Tupi perfectly well from the beginning:


For example, on the very day of his capture he explained (Staden 1928:65): “The savages asked me whether their enemies the Tuppin Ikins had been there that year to take the birds during the nesting season. I told them [emphasis added] that the Tuppin Ikins had been there, but they proposed to visit the island to see for themselves . . . ” If Staden did not speak Tupi at the time of his capture, then there is no way that he could have told them anything, since it is hardly likely that his captors spoke German or Portuguese (ibid., 21).


It is in fact very probable that Staden had learnt Tupi well before his capture, since he had lived on the coast of Brazil for two years with a number of other Europeans before he fell into the hands of the Tupinamba. During this time there was constant contact with local Indians who spoke the Tupi language, which was common to a number of tribes besides the Tupinamba. As Forsyth says, “Tupi was the lingua franca of Brazil at this time (and for a long time to come). The Europeans learned to speak Tupi, rather than the Tupians learning French or Portuguese” (ibid., 22-3). After a year, Staden and other Europeans reached the Portuguese settlement of Sao Vicente. He worked for the Portuguese for a year, during which he was given a Tupi-speaking slave who worked for him on a daily basis, which itself gave him ample opportunity to learn the language.


Arens is also entirely mistaken when he claims that the Tupi would have had to understand German when responding to Staden’s singing of a psalm.


. . . this is simply not so. Staden (1928:67) actually says: “So in mighty fear and terror I bethought me of matters which I had never dwelt upon before, and considered with myself how dark is the vale of sorrows in which we have our being. Then, weeping, I began in the bitterness of my heart to sing the Psalm: ‘Out of the depths have I cried unto thee.’ Whereupon the savages rejoiced and said: ‘See how he cries: Now he is sorrowful indeed’ [emphasis added]”. It is not to the German words of the psalm that the Indians respond, rather to the fact that Staden was weeping (Forsyth 1985:23-4).


Arens also refers to


. . . a small paragraph which curiously informs the reader that “the savages have not the art of counting beyond five” . . . Consequently, they often have to resort to their fingers and toes. In those instances when higher mathematics are involved extra hands and feet are called in to assist in the enumeration. What the author is attempting to convey in his simple way with this addendum is that the Tupinamba lack culture in the sense of basic intellectual abilities. The inability to count is to him supportive documentation for the idea that these savages would resort to cannibalism. To Staden and many others, eating human flesh implies an animal nature which would be accompanied by the absence of other traits of “real” human beings who have a monopoly on culture (Arens 1979: 23-4).


Chagnon (1977: 74) states that the Yanomamo only have words for one and two, and I record that the same is true of the Tauade; neither of us, however, was trying to insinuate that the Yanomamo and the Tauade were therefore subhuman animals, and Forsyth adds that “Arens completely ignores the fact that Staden’s statement concerning Tupinamba enumeration is correct. Ancient Tupi had no terms for numbers beyond four. Larger numbers were expressed in circumlocutions, often involving fingers and toes” (Forsyth 1985: 19). If Arens were better informed he would know that very restricted number systems are often found among hunter-gatherers and simple cultivators, and this condescending, ad hominem attack on Staden tells us much more about Arens’s prejudices than about Staden’s.


Finally, Arens tries to argue that later authors who at first sight appear to confirm Staden’s account of cannibalistic ceremonies were in fact simply plagiarising him. Forsyth, however, dismisses the claim of plagiarism entirely:


Arens’s (1979: 28-30) whole argument is based on the similarities in the accounts of Staden, Lery (1974: 196), Thevet (1971: 61-63), Knivet (1906: 222), and Casas (1971: 68) with respect to the verbal exchange between the victim and executioner before an enemy was killed, cooked, and eaten. His argument is as follows:


In his chapter on killing and eating the victim, Staden supplies some further Indian dialogue which he translates for his readers. He states that the Indian who is about to slay the prisoner says to him: “I am he that will kill you, since you and yours have slain and eaten many of my friends.” The prisoner replies: “When I am dead I shall still have many to avenge my death” [Staden 1928: 161]. Dismissing the linguistic barrier momentarily . . . the presentation of the actual words of the characters lends an aura of authenticity to the events. However, if similar phrases begin make their appearance in the accounts of others who put themselves forward eyewitnesses to similar deeds, then the credibility of the confirmation process diminishes (Arens 1979: 28-29). Arens cites the other authors to show the similar phraseology used in describing the execution scene. Hence his whole case for plagiarism is similarities in two sentences in works that are book length in most instances (see Riviere 1980: 204).


As it turns out, however, when even these two sentences are examined in the context of what we know about the cannibalistic rites themselves, and about how and when the accounts were produced, Arens’s argument evaporates. An example from our own literate society should suffice to show why this is so. If several different observers wrote a description of the Pledge of Allegiance ceremony, which takes place daily in schools all over the nation, we should hardly be surprised to find considerable similarity, since what is said is an essential element in the ceremony. But according to Arens’s logic, we would have to conclude that the writers were all copying one another. But the Pledge of Allegiance is not a random event in the daily activities of American school-children. It is, rather, a ritual charged with symbolic meaning. In such a ritual the repetition of behavior and utterance is an integral part of the ceremony . . . The verbal exchanges cited by Arens between executioner and victim were not simply random babblings, but highly ritualized exchanges constrained by custom and belief at the very climax of the ceremony, as virtually all of the accounts make patently clear . . . (Forsyth 1985: 27-8).


Forsyth also points out that Arens ignores a wealth of Jesuit sources that provide eye-witness accounts of cannibalism, the confiscation of cooked (and preserved) human flesh from the Indians (so that they would not eat it), the confiscation of bodies from Indians who were about to eat them, or persuading them to bury the bodies rather than eating them—in one case after the body was already roasted and, in another, the successful rescue of prisoners before they could be killed and eaten:


Whatever the reliability of the better-known sources may be, the Jesuit sources are unimpeachable in this matter, because they avoid all of the alleged weaknesses of the accounts referred to by Arens. They are not copies of Staden, Lery, or Thevet; many of the letters and reports were written before these authors even arrived in Brazil. Moreover, many of the Jesuits did speak the Tupi tongue, even writing dictionaries and grammars to help others learn the language, and lived in Indian villages for extended periods of time. In addition, details of the various Jesuit accounts often differ sufficiently from one another to rule out plagiarism (Forsyth 1983: 171).


Just as Forsyth claims that Arens ignores a wide range of original sources, particularly those of the Jesuits, Neil Whitehead (1984) also documents Arens’s similar failure to consult Jesuit sources with regard to the separate issue of Carib cannibalism.


So far, we have been considering accounts of cannibalism that involve the eating of enemy prisoners, usually killed or captured in warfare. Cross-culturally this appears to be the basic form of cannibalism; there seems little evidence that shortage of protein had anything to do with it, as materialists like Marvin Harris supposed; and many primitive societies were as strongly opposed to cannibalism as we are. There is, however, a different type of cannibalism, conventionally known as “endo-cannibalism,” in which the relatives of a deceased person eat the corpse, or part of it, as a mortuary rite. Roy Wagner gives a detailed account of this among the Daribi of the New Guinea Highlands (1967: 145-7), and to a very limited extent the Tauade also practised this:


When a person died and his body had rotted in the tseetsi [a raised basket] or in the ground, the bones were taken by his relatives and washed in a stream. The skull in particular was washed out with water introduced through the foramen magnum, with which the remains of the brain were flushed away. The children of the deceased are said to have drunk this water (Hallpike 1977: 158).


Arens, however, is forced to be just as dismissive of endo-cannibalism as he is of cannibalism in general and occupies many pages in particular trying to discredit the accounts of this practice among the Fore of New Guinea, which became world-famous through its association with two Nobel Prize winners. Most people would probably consider the Fore case a major obstacle to his theory, and Arens’s attempts to dismiss it are excellent examples of the quality of his research. Patrol reports in the Fore area from the early 1950s onwards began describing a disease that became known as kuru. Its symptoms were trembling, difficulty in walking and co-ordination, mood changes, and slurred speech, leading to unconsciousness and death usually within a year or less from the first symptoms appearing. (The word kuru itself referred to the casuarina tree, whose quivering leaves were seen by the Fore as similar to that of the victims’ limbs.) The American physician Carleton Gajdusek happened to be in the area and was told about the disease by Dr Vincent Zigas. The anthropologist Ronald Berndt had already studied it, and considered it psycho-somatic, but Gajdusek came to the firm conclusion that it was entirely physical in origin, and in 1957 Gajdusek and Zigas published a paper claiming that it was a newly discovered neurological disease.


Initially it had been supposed that it might be genetic in origin, but this would have required a long evolutionary history and resulted in epidemiological equilibrium, whereas the Fore claimed that it had first appeared around the beginning of the century, thirty years before contact with Europeans, and its incidence had steadily increased throughout the 1940s and 1950s and was now killing very significant numbers of people. The mortality rate in some villages was 35/1000 per annum, and far more women than men were affected (Lipersky 2013: 479). In 1957, for example, approximately 170 women died compared to 35 men (ibid., Fig. 4, 476). In 1961 the anthropologists Robert and Shirley Glasse (later Lindenbaum) carried out fieldwork on the Fore, with the specific purpose of seeing if victims were close relatives, as the genetic hypothesis predicted, but discovered that they were not. They also made special enquiries into the endo-cannibalistic practices of the Fore, which had been suppressed some years before their work. In the late 1930s and 1940s, many gold miners, Protestant missionaries, and government officials (in other words, Arens’s usual “basket of deplorables” in this scenario), had already become familiar with the presence of endo-cannibalism among Eastern Highland tribes (Lipersky 2013: 475). The Glasses made their own enquiries from informants and were able to reconstruct the ways in which this had been carried out on the body of a deceased relative:


When a body was considered for human consumption, none of it was discarded except the bitter gall bladder. In the deceased’s old sugarcane garden, maternal kin dismembered the corpse with a bamboo knife and stone axe. They first removed hands and feet, then cut open the arms and legs to strip the muscles. Opening the chest and belly, they avoided rupturing the gall bladder, whose bitter content would ruin the meat. After severing the head, they fractured the skull to remove the brain. Meat, viscera, and brain were all eaten. Marrow was sucked from cracked bones, and sometimes the pulverized bones themselves were cooked and eaten with green vegetables. (Lindenbaum 2013:224)


They also found that


Some elderly men rarely ate human flesh, and small children residing with their mothers ate what their mothers gave them. Youths, who were initiated around the age of ten, moved to the men’s house, where they began to observe the cultural practices and dietary taboos that defined masculinity. Consuming the dead was considered appropriate for adult women but not men, who feared the pollution and physical depletion associated with eating a corpse. The epidemiological information provided by Gajdusek and Zigas in 1957—that kuru occurred among women, children of both sexes, and a few elderly men—seemed to match perfectly the Fore rules for human consumption (Lindenbaum 2015: 104-5).


Which it did. To cut a long story short, Gajdusek was joined in his research by Stanley Prusiner, a biologist who, like Gajdusek, received the Nobel Prize. The genetic basis of kuru had been rejected, and Gajdusek had shown that the disease could be transmitted to primates exposed to infected material, which suggested to him that the disease was carried by a slow virus. Prusiner, however, showed that kuru was actually caused by prions, an abnormal type of protein, which contain no genetic material, and was a spongiform encephalitis in the same family as Creutzfeldt-Jakob disease. The point about prions was that, whereas a slow virus would allow kuru to be spread simply by contact, prions required the actual consumption of brain matter, and the obvious occasions for this were the Fore mortuary ceremonies in which the women ate the brains of the deceased. With the demise of cannibalism the incidence of kuru fell steadily over the years, and by 1982 there were very few deaths, and the sex ratios were now equal (Lipersky 2013: Fig. 4, 476). The disease is currently considered extinct.


While Arens admitted that ‘it is impossible to prove that cannibalism is not a factor in the kuru syndrome’, he nevertheless was not convinced: the evidence was circumstantial, there were contradictions in the ethnography, and the same material lent itself to alternate explanations (Arens 1979: 112). He points out that Fore cannibalism had never been observed by an outsider, and that the anthropologists were uncertain when it had been abolished. “As a result, Glasse and Lindenbaum relied upon Berndt’s idiosyncratic discussion of the material, the fact that the Fore had a reputation among surrounding groups for eating their dead, the odd report that someone had eaten someone else and the belief among the males that ‘the great majority of women were cannibals’ (109).” Of this belief about the Fore women he says: “Rather than uncritically accepting the native view that only women and children are cannibals, it would seem reasonable to question whether or not this might be a symbolic statement about females, in a culture area renowned for sexual antagonism and opposition . . . ” (110). He goes on, “Another reasonable suspicion of the cannibalism hypothesis is aroused by the fact that among the Fore each death is followed by a mortuary feast involving the slaughter of pigs and distribution of the meat and vegetables . . . This period of an abundance of animal protein would seem to be the least likely time to resort to cannibalism” (111).


With regard to the transmission of the disease, which by 1979 had been accepted as related to the Creutzfeldt-Jakob family, he remarks that no one has suggested that such diseases “are transmitted in the western world by cannibalism. However, such a hypothesis presents no problem when the affected population is the inhabitants of the New Guinea highlands. This is consistent with the general theoretical tone of much of the anthropological literature on this area, which effectively diminishes the cultural achievements of the inhabitants” (112). With regard to the initial appearance of the disease he says, “Surprisingly enough, no one has seriously considered the idea that the presence of Europeans in the area was responsible for the outbreak of the epidemic at the turn of the century. The arrival of the first two Europeans in 1932 does not deny the possible entry of the disease years before through indirect means and intermediaries” (113). He also points out the important social changes that have occurred since European contact, such as the disuse of the men’s house and men moving into live with their wives and children: “In the light of the obvious cultural rearrangements and new experiences, it is odd that scientific researchers have seized on a correlation between something which was never seen and another phenomenon studied and measured so meticulously” (113).


Arens’s hilarity at the racist idea of Creutzfeld-Jacob disease being transmitted by cannibalism turned out to be misplaced, however, since it was cattle cannibalism in the form of brain and spinal cord matter from diseased animals being included in cattle feed that led, a few years later, to the spread of BSE in Britain. Bovine Spongiform Encephalopathy, or Mad Cow Disease, was a prion disease that also infected a number of humans in the form of vCJD, variant Creutztfeldt-Jakob Disease, as a result of eating this meat and led to a ban on the export of British beef in 1996.


In 1997, in “Man is off the menu,” he added a further refinement to his “refutation” of Fore cannibalism, which is worth quoting as an example of his methods of scholarly disputation:


There was a particularly notable agreement [among anthropologists] that cannibals did exist, however, until practically yesterday, in the highlands of New Guinea, the “final frontier” of western cultural contact. In this instance many smugly noted that the evidence for cannibalism emerged from medical research rather than from the usual less reliable forms of documentation. In the light of the exalted position of science, how could any rational person doubt this research? I discovered, with perhaps even more smugness, that one could. The story began in 1957, with the arrival in New Guinea of D. Carleton Gajdusek, an American research paediatrician on his way home from a fellowship year in Australia. Why he opted to visit this part of the world did not become clear until recently. However, the eventual results of the sojourn proved important for both medical science and for Dr Gajdusek. Eventually, he would receive the Nobel prize for medicine, and then, later, be arrested and plead guilty to the sexual abuse of minors in the US. He adopted a number of boys from part of New Guinea well known for institutionalised male homosexuality between youngsters and adults. Laudatory reports of Gajdusek’s charity, including references to his bringing a number of the lads to the Nobel ceremonies, were recounted in the media (Arens 1997:16).


Gajdusek’s subsequent criminal conviction related to boys of a different people from the Fore and had nothing whatever to do with his kuru research, and therefore provides Arens with no grounds for doubting it, smugly or otherwise. Arens, of course, as we might expect, makes no reference in his article to Prusiner’s work and the crucial association of brain-matter with prions which was conclusive support for the cannibalistic thesis, and by 1997 had been well-established.


I leave it to my readers to decide if they find these various arguments of Arens even a remotely adequate response to the facts presented on Fore cannibalism. Shirley Lindenbaum comments that “Although discredited today, the denial of cannibalism was kept alive during the 1980s and 1990s by a generational shift in the human sciences, glossed as postmodernism, which studied metaphor and representation, providing new life for the idea that cannibalism was nothing more than a colonizing trope and stratagem, a calumny used by colonizers to justify their predatory behaviour” (Lindenbaum 2015: 108).


To sum up, then, Arens’s charge that anthropologists engage in “manipulating the data to generate a foregone conclusion” where “academic standards seem to function as an almost forgotten ideal,” actually turns out to be a very accurate description of his own book, and Marshall Sahlins, who has done more than most to refute it, may be allowed the last word:


It all follows a familiar American pattern of enterprising social science journalism: Professor X puts out some outrageous theory, such as the Nazis really didn’t kill the Jews, human civilization comes from another planet, or there is no such thing as cannibalism. Since facts are plainly against him, X’s main argument consists of the expression, in the highest moral tones, of his own disregard for all available evidence to the contrary. He rises instead to the more elevated analytical plane of ad hominem attack on the authors of the primary sources and those credulous enough to believe them. All this provokes Y and Z to issue a rejoinder, such as this one. X now becomes “the controversial Professor X” and his book is respectfully reviewed by non-professionals in Time, Newsweek, and The New Yorker. There follow appearances on radio, TV, and in the columns of the daily newspapers (Sahlins 1979).



1. ^ The class of stupid, ignorant people.


2. ^ For Cook’s actual Journal entry see J.C. Beaglehole, ed., 1969. The Voyage of the Resolution and Adventure 1772-1775 (Cambridge: The University Press for the Hakluyt Society), pp. 292-293.


3. ^ But Sahlins also explains that the authorship of this account might have been mistakenly attributed to Endicott:

It could be that Endicott indeed did not see the event, insofar as he may well not be the author of the contested text. The original of that text, reprinted and signed by Endicott as an appendix to his book, is an article that appeared in The Danvers Courier newspaper on 16 August 1845, under the byline ‘By an Eye Witness’. The Peabody Museum, where the article is archived, apparently attributes it to a different member of the Glide’s crew, Henry Fowler (of Danvers) with whose papers it is included (Fowler, PMB 225). Indeed, a simple ‘F’ is inscribed at the bottom of the original newspaper article (Sahlins 2003: 3, n.3)  But whether Endicott or Fowler provided the actual account, it is confirmed by other members of the Glide’s crew.




Arens, W. 1979. The Man-Eating Myth. Anthropology & anthropophagy. Oxford University Press.


Arens, W. 1997, ‘Man is off the menu’, Times Higher Education, 1310, 16.


Arens, W. 2003. ‘Cannibalism reconsidered’, Anthropology Today, 19(5), 18-19.


Beaglehole, J.C. 1974. The Life of Captain James Cook. Stanford University Press.


Brown, P., & Tuzin, D. (eds.) 1983. The Ethnography of Cannibalism. Washington: The Society for Psychological Anthropology.


Endicott, W. 1923. Wrecked Among Cannibals in the Fijis. Salem, Massachusetts: Marine Research Society.


Forsyth, D. 1983. ‘The beginnings of Brazilian anthropology: Jesuits and Tupinamba cannibalism’, Journal of Anthropological Research, 39(2), 147-78.


Forsyth, D. 1985. ‘Three cheers for Hans Staden: the case for Brazilian cannibalism’, Ethnohistory, 32(1), 17-36.


Hallpike, C.R. 1977. Bloodshed and Vengeance in the Papuan Mountains. The generation of conflict in Tauade society. Oxford: Clarendon Press.


Hallpike, C.R. 2008. The Konso of Ethiopia. A study of the values of an East-Cushitic people. 2nd ed. Bloomington, Indiana: AuthorHouse.


Jennings, W. 2004. ‘The debate over kai tangata (Maori cannibalism): new perspectives from the correspondence of the Marists’, Journal of the Polynesian Society, 120(2), 129-47.


Koch, K-F. 1974. War and Peace in Jalemo. The management of conflict in Highland New Guinea. Harvard University Press.


Lindenbaum, S. 2013. Kuru Sorcery. Disease and Danger in the New Guinea Highlands. 2nd Ed. Colorado: Paradigm Publishers.


Lindenbaum, S. 2015. ‘An annotated history of Kuru’, Medical Anthropology Theory, 2(1), 95-126


Liperski, P.P. 2013. ‘Kuru: a journey back in time from Papua New Guinea to the Neanderthals’ extinction’, Pathogens, 2(3), 472-505.


Martin, J. 1827. An Account of the Natives of the Tonga Islands. 2 vols. 3rd Ed. Edinburgh: Constable & Co.


Prusiner S.B., Gajdusek D.C., Alpers M.P. 1982. ‘Kuru with incubation periods exceeding two decades’,  Ann. Neurol. 12


Sahlins, M. 1979. ‘Cannibalism: An Exchange’, with W.Arens, New York Review of Books, March 22.


Sahlins, M. 1983. ‘Raw women, cooked men, and other “Great Things” of the Fiji islands’, in Brown & Tuzin (eds), The Ethnography of Cannibalism, 72-93, Washington: Society for Psychological Anthropology.


Sahlins, M. 2003. Artificially maintained controversies. Global warming and Fijian cannibalism’, Anthropology Today, 19(6), 3-5.


Wagner, R. 1967. The Curse of Souw. Chicago University Press.


Whitehead, N.L. 1984. ‘Carib cannibalism: the historical evidence’, Journal de la Société des Américanistes, 70, 69-87




C. R. Hallpike is Emeritus Professor of Anthropology at McMaster University, Canada, and an Oxford D.Litt, and spent a lifetime’s research living with mountain tribes in Ethiopia and Papua New Guinea and writing many books about them and on morality, religion, culture and social evolution. He is the author of Do We Need God To Be Good? (2017), Ethical Thought in Increasingly Complex Societies: Social Structure and Moral Development (2016), On Primitive Society: and other forbidden topics (2011), and How We Got Here: Bows and Arrows to the Space Age (2008).

NER on Twitter @NERIconoclast


Leave a Reply

Your email address will not be published. Required fields are marked *

New English Review Press is a priceless cultural institution.
                              — Bruce Bawer

Order here or wherever books are sold.

The perfect gift for the history lover in your life. Order on Amazon US, Amazon UK or wherever books are sold.

Order on Amazon, Amazon UK, or wherever books are sold.

Order on Amazon, Amazon UK or wherever books are sold.

Order on Amazon or Amazon UK or wherever books are sold

Order at Amazon, Amazon UK, or wherever books are sold. 

Order at Amazon US, Amazon UK or wherever books are sold.

Available at Amazon US, Amazon UK or wherever books are sold.

Send this to a friend