You are on page 1of 9

On the Definition of Shamanism Author(s): PieterJolly/CeceliaF.Klein, EulogioGuzmn, and MayaStanfieldMazzi Source: Current Anthropology, Vol. 46, No.

1 (February 2005), pp. 127-128 Published by: The University of Chicago Press on behalf of Wenner-Gren Foundation for Anthropological Research Stable URL: http://www.jstor.org/stable/10.1086/427098 . Accessed: 29/11/2013 20:20
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

The University of Chicago Press and Wenner-Gren Foundation for Anthropological Research are collaborating with JSTOR to digitize, preserve and extend access to Current Anthropology.

http://www.jstor.org

This content downloaded from 200.26.133.57 on Fri, 29 Nov 2013 20:20:43 PM All use subject to JSTOR Terms and Conditions

Discussion

On Modern Behavior and the Evolution of Human Intelligence


john r. f. bower 3496 Oyster Bay Ave., Davis, CA 95616, U.S.A. (jrfbower@aol.com) 20 vii 04 Henshilwood and Marean (CA 44:62751) have opened a number of useful perspectives on the role of modern behavior in the late stages of human evolution. Salient among these is a critique of the integrated trait-list approach to identifying modern behavior, wherein such behavior is dened in terms of a presumed more or less simultaneous earliest occurrence of specic technological capabilities, together with the use of a wide variety of materials (bone, antler, shell, etc.) for fabricating tools and other artifacts, advanced predatory capabilities, production of artifacts (such as body ornaments) whose main purpose is symbolic, and so on. As they point out, the trait list appears to be disintegrating as a result of discoveries, particularly in Africa, indicating that the earliest appearance of various items in the list is spread over a span of time measured in tens of thousands of years (Brooks et al. 2004, Henshilwood et al. 2004, Holden 2004, Thompson et al. 2004). Although I applaud Henshilwood and Mareans deconstruction of the trait-list approach to archaeological investigation of modern behavior, I think we need to beware of throwing the baby out with the bathwater. The items that make up the trait list do, in fact, coalesce in Paleolithic site inventories somewhere in the range of 40,00050,000 years ago, signaling a major inection in humanitys global history broadly parallel to other revolutions (Neolithic, Industrial, etc.) that our species has experienced, and therefore we need to ask what precipitated this punctuated episode in Paleolithic culture history. As Henshilwood and Marean point out, the list of possible contributing factors includes language, symbolically organized behavior, and demographic pressure. However, all of these (and all other less thoroughly scrutinized prospects mentioned in the article) are problematic. For example, language is not reected in the material remains of preliterate cultures and, in any case, may have evolved over a huge expanse of time perhaps dating back to the earliest appearance of the genus Homo ca 2.5 million years ago (Jerison 2001), which renders it poorly suited to accounting for punctuated development. Symbolically organized behavior is conceptually vague and, partly for this
Permission to reprint items in this section may be obtained only from their authors.

reason, difcult to identify in the material remains of Paleolithic culture. And demographic pressure is unlikely to have been widely experienced by Paleolithic cultures during the relevant time frame. Quite apart from such specic problems, there are deeper aws in the prevailing archaeological program for discovering the roots of modern behavior. First, as recognized by Henshilwood and Marean and emphasized in a comment on their article by Chase, the term modern behavior is utterly inappropriate in the context of Paleolithic culture. What is modern today will not be tomorrow, was not yesterday, and certainly was not 40,000 50,000 years ago. In addition, there is an underlying assumption that modern behavior (or whatever we call it) ultimately resulted from the evolution of human intelligence. Thus, for example: The fundamental adaptation of modern humans is culture and technology, and both are heavily conditioned by intellect (p. 645). This leads to a quest for a Paleolithic IQ test (or tests) that will reveal the degree and kind of intelligence that sustains fully symbolic sapiens behavior or whatever synonym we prefer. I believe it may prove more efcacious to set aside the concept of intelligence and focus more narrowly on specic cognitive attributes not how well Paleolithic people thought but the content of their thoughts (Wynn 2001). This not only narrows the eld of inquiry but allows us to examine attributes that are (a) archaeologically visible, (b) tightly coupled with basic, universal features of human behavior, and (c) susceptible to neurological and even possibly genetic explanation. Such an approach may also allow us to escape the circularity that stems from seeking to explain the origin of one kind of behavior (e.g., modern) in terms of another form of behavior (e.g., technological prociency). Self-awareness, or self-recognition, is a cognitive attribute that is archaeologically attested in a reasonably straightforward way through artifacts that served as human body ornaments (beads, pendants, etc.). It is also an attribute that is crucially linked to fundamental aspects of human psychology and social behavior. This includes the formation of human groups (it is no accident that kinship charts are anchored on ego), a preoccupation with mortality, many (if not all) human emotional states, and even consciousness. As Metzinger (2003) has argued, the self is a mental construct that not only gives rise to subjective experience but also is the foundation for such human capabilities as abstract thinking and understanding the minds of others. Moreover, progress is being made in research aimed at revealing the neurological basis for the experience of self (e.g., Ehrsson, Spence, and Passingham 2004), which may eventually open the door to discovering its genetic foundation. A line of inquiry aimed at understanding the evolution 121

This content downloaded from 200.26.133.57 on Fri, 29 Nov 2013 20:20:43 PM All use subject to JSTOR Terms and Conditions

122 F c u r r e n t a n t h ro p o l o g y

of the modern human mind through the attribute of selfrecognition may lead into a labyrinth of interdisciplinary investigation. This is true of cognitive archaeology in general (Wynn 2001, Chase 2001); it is both the challenge of the eld and its reward.

References Cited
b ro o k s , a . j . , j . f e a t h e r s , g . h a r t m a n , c . t ru e m a n , n . t u r o s s , f . d e r r i c o , a n d j . y e l l e n . 2004. Middle Stone Age barbed bone points from Katanda (DR Congo): New perspectives on age and association. Paper presented at the annual meeting of the Paleoanthropology Society, Montreal, March 3031. c h a s e , p . 2001. Multilevel information processing, archaeology, and evolution, in In the minds eye. Edited by April Nowell, pp. 12135. Ann Arbor: International Monographs in Prehistory. ehrsson, h. h., h. c. spence, and r. e. pass i n g h a m . 2004. Thats my hand! Activity in premotor cortex reects feeling of ownership of a limb. Science Express, July 1, pp. 14. http://www.scienceexpress.org. h e n s h i l w o o d , c . s . , a n d c . m a r e a n . 2003. The origin of modern human behavior: Critique of the models and their test implications. current anthropology 44:62751. henshilwood, c. s., f. derrico, m. vanhaeren, k . v a n n i e k e r k , a n d z . j a c o b s . 2004. Middle Stone Age shell beads from South Africa. Science 304:404. h o l d e n , c . 2004. Oldest beads suggest early symbolic behavior. Science 304:369. j e r i s o n , h . j . 2001. Archaeological implications of paleoneurology, in In the minds eye. Edited by April Nowell, pp. 8396. Ann Arbor: International Monographs in Prehistory. m e t z i n g e r , t . 2003. Being no one: The self-model theory of subjectivity. Cambridge: MIT Press. thompson, j., j. bower, e. sher, a. mabulla, c. m a r e a n , k . s t e w a r t , a n d c . v o n d r a . 2004. Loiyangalani: Behavioral and taphonomic aspects of a Middle Stone Age site in the Serengeti Plain, Tanzania. Paper presented at the annual meeting of the Paleoanthropology Society, Montreal, March 3031. w y n n , t . 2001. The role of archaeology in cognitive science, in In the minds eye. Edited by April Nowell, pp. 919. Ann Arbor: International Monographs in Prehistory.

On Breast Milk, Diet, and Large Human Brains


josephine c. a. joordens, remko s. kuipers, and f r i t s a . j . m u s k i e t Brouwersgracht 120, 1013 HA Amsterdam, The Netherlands (joordens.dogger@worldonline.nl)/ Pathology and Laboratory Medicine, University Hospital Groningen, P.O. Box 30.001, 9700 RB Groningen, The Netherlands (f.a.j.muskiet@lc.azg.nl). 19 vii 04 Robson, in Breast Milk, Diet, and Large Human Brains (CA 45:41925), has tested the hypothesis that greater

proportions of animal products in human diet affect the composition of human milk and Martins hypothesis that the unique pattern of human brain growth is due to the unique composition of human milk. Her conclusions are that the wide variation in human diet has little impact on breast milk quantity and quality and that Martins hypothesis should be rejected. We agree with the latter conclusion (adding that we question whether Martins hypothesis merited the effort) but would like to raise some points of discussion. 1. Robson states that AA and DHA levels in human milk are consistent despite variations in maternal diet. Contrary to this, however, she notes the dose-dependency of its DHA content indicated by several maternal-DHA or sh-oil-supplementation trials. Consequently, because of the sizable worldwide differences in maternal DHA (especially sh) intake, one may expect the DHA content of human milk to be subject to wide biological variation, and that is indeed what we have found in a data set of 465 human milk samples collected by the Groningen University Hospital in various countries of the world during the past 25 years (Smit et al. 2002). Analysis of 28 fatty acids in these samples revealed that the biological variation is greatest in DHA and EPA, amounting to 68% and 100%, respectively. The lowest DHA content was found in the breast milk of a mother living in Islamabad, northern Pakistan (0.03 mol%), and the highest in a counterpart living in one of the sh-eating Caribbean islands (1.63. mol%). The low DHA content of milk in northern Pakistan is a matter of special concern (Smit et al. 1999), since it undoubtedly reects poor maternal DHA status in pregnancy, during which the initial part of brain growth occurs. Regarding the consistency of the AA content of human milk Robson seems to have a point, since in line with this notion our data set indicates relatively low (i.e., 28%) biological variation. However, as yet unpublished data obtained from a sh-eating population living on Lake Kitangiri (a freshwater lake in northeastern Tanzania) has refuted this supposed consistency. The milk of these women was found to have the highest AA content encountered by us so far (median 0.70 mol%, range 0.500.90), which derives from the lifetime consumption of local AA- (and DHA-) rich freshwater sh as the only animal lipid source. We conclude that the fatty-acid composition of human milk is highly dependent on both short- and long-term maternal diet and that, in view of the variety of dietary habits, there is no such thing as consistency in its fatty-acid composition worldwide. We have at present little knowledge of the evolutionarily established unique composition of human milk or the interaction between this ancient human milk composition and our brain-building genome. It is, however very probable that early Homo sapiens had higher intakes of longchain polyunsaturated fatty acids (LC-PUFA) (e.g., from game and sh) than modern Western humans, and recent experiments in rats have shown that dietary omega-3 fatty acids (ALA, EPA, and DHA) not only cause changes in brain phospholipid structure and

This content downloaded from 200.26.133.57 on Fri, 29 Nov 2013 20:20:43 PM All use subject to JSTOR Terms and Conditions

Volume 46, Number 1, February 2005 F 123

composition but also modulate the expression of a sizable number of genes with functions in, for example, brain cell structure, energy metabolism, neurotransmission, signal transduction, and regulation (Kitajka et al. 2002). 2. Robson concludes that there is no evidence that LC-PUFA deciencies during infancy cause smaller brain size. We would point out that many studies have shown a relation between DHA status and head circumference or calculated brain weight, though only within the range of normal human brain proportions (e.g., Hornstra 2000, Woltil et al. 1998). However, we regard brain size as irrelevant to the issue at hand. It is not brain size that matters here but brain quality. Therefore, we do not agree with Robsons support of Heirds observation that formulas without LC-PUFA have been fed to infants for decades and have not resulted in epidemics of either poor vision or neurodevelopmental delays. It is well established (see, e.g., Muskiet et al. 2004) that feeding infants with standard formulas lacking DHA and AA causes biochemically demonstrable low status of LC-PUFA in various body compartments including the brain (notably DHA). Low LC-PUFA status coincides, especially in preterm infants, with unfavorable neurodevelopment during the rst four postnatal months, as derived from various tests of visual, perceptive, cognitive, and motor development (e.g., Larque, Demmelmair, and Koletzko 2002, Koletzko et al. 2001, Innis 2003). A recent study carried out by our Groningen group showed that term infants receiving formula without LC-PUFA had a higher frequency of mildly abnormal general movements at the age of three months, a condition which signicantly increased risk of development of minor neurological dysfunction, attention problems, and aggressive behavior at school age (Bouwstra et al. 2003). In addition, epidemiological data and controlled trials with ALA and sh oil implicate subclinical omega-3 deciency in cardiovascular disease, inammatory disorders, attention decit disorder (ADHD), dyslexia, (postpartum) depression, dementia, and schizophrenia (Muskiet et al. 2004 and references therein, Peet 2004). Corresponding with declining intake of omega-3 fatty acids during the past 100 years, some of these diseases have become widespread in the modern Western world and nowadays constitute a major public health concern. We realize that the consequences of subclinical (micro-) nutrient deciencies are difcult to objectify and to prove according to rules of evidence-based medicine, since deciencies usually operate over long periods, which hampers identication of causal relationships. However, we feel that the current lack of hard evidence on long-term effects of low LC-PUFA status in pregnancy and infancy is not to be taken as proof of lack of effect and therefore regret the misleading message transmitted by the quotation from Heird in Robsons report.

Reply
s h a n n e n l . ro b s o n Department of Anthropology, 270 South 1400 East, Room 102, University of Utah, Salt Lake City, UT 84112, U.S.A. (robson@umnh.utah.edu). 1 x 04 I thank Joordens, Kuipers, and Muskiet for their comments and appreciate the opportunity to continue discussion surrounding the mechanisms of human encephalization. While they contest the generalizations I make about the cross-cultural variation of human breast milk composition and its effects on infant development, they agree with my conclusion that human milk is similar to that of other primates and is therefore not a unique resource supporting the unique rapid postnatal brain growth in humans as Martin hypothesized. However, they question whether testing Martins hypothesis merited the effort. Both Martins hypothesis and meat eating as a prime releaser for large human brain size remain extremely inuential propositions among anthropologists even in the absence of a pathway for diet to facilitate this expansion. If meat eating is to inuence human brain growth, then it must be shown what biochemical properties affect brain growth, when ontogenetically these nutrients would be inuential, and how meat eating would transfer these benets. I addressed each of these lines of evidence and tested Martins (1983, 1995) hypothesis that human breast milk was unique in composition, conferring nutritional benets to infants during the period of rapid postnatal brain growth. I showed that, contrary to long-standing assumptions about the role of meat eating and breast milk composition in human encephalization, the long-chain fatty-acid (LC-PUFA) composition of human breast milk is not unique among anthropoid primates and that the two primary fatty acids in brain tissue, AA and DHA, are not prime releasers for encephalization. Joordens et al. cite two studies to contend that brain size is positively correlated with DHA levels. However, neither Hornstra (2000) nor Woltil et al. (1998) show that full-term infants fed non-LC-PUFA formula milks have smaller adult brain sizes. Hornstra (2000) investigated the DHA levels of fetal tissue, not breast-feeding infants, and did not nd a statistically signicant correlation with smaller head circumference after correction for gestational age (p. 1264S). Woltil et al. (1998) studied low-birth-weight babies, over half of whom were premature. Low birth weight and prematurity are both variables known to have important confounding effects on infant outcome (Hack et al. 2002). Nevertheless, as Joordens et al. note, the brain weights of infants in both studies were within normal ranges, further underlining the lack of evidence linking infant LC-PUFA levels and resulting brain size. Joordens et al. say that they regard brain size as irrelevant to the issue at hand. It is not brain size that matters here but brain quality. While brain size was exactly the issue at hand in my report, they focus on brain quality

This content downloaded from 200.26.133.57 on Fri, 29 Nov 2013 20:20:43 PM All use subject to JSTOR Terms and Conditions

124 F c u r r e n t a n t h ro p o l o g y

because they have two primary objections to generalizations I make about the consistency of fatty acids in human milk and the (supercial) effects of LC-PUFA during infant development. First, they cite the analyses of Smit et al. (2002) and unpublished data showing wide variation in the fatty-acid levels of human milk and no such thing as consistency in its fatty-acid composition worldwide. Smit et al. (2002) qualify the wide variation in their study as somewhat articial (p. 551) because a few cases of very high DHA levels skew their distribution. In the data they used (table 1, p. 552) the populations range lows (0.080.10 mol%) and medians (0.160.33 mol%) are remarkably similar despite very different diets. DHA is always present in breast milk, regardless of maternal diet, and the median value of DHA in human milk is, in fact, very consistent cross-culturally. Second, Joordens et al. point to numerous studies showing negative health and neurological effects coincident with low LC-PUFA status and therefore nd Heirds (1999) conclusion that the described benets of LC-PUFA are subtle and that formulas without LC-PUFA have been fed to infants for decades and have not resulted in epidemics of either poor vision or neurodevelopmental delays (p. 207) misleading. Though they note (as do I) a lack of hard evidence supporting the long-term effects of low LC-PUFA status, they argue that this should not be taken as proof of lack of effect. Innis (2003) recently reviewed the mixed results of studies exploring a relationship between DHA status and several mental or motor development tests. The available evidence is equivocal, and Innis reminds us that those studies which do nd associations between DHA and visual and neurodevelopment in breast-fed infants should not be confused with demonstration of causality (p. S5). Recently, Muskiet et al. (2004) continued to draw attention to the improbability of linking meat eating, LCPUFA dietary intake, and evolution of large human brain size. They agree that meat is a poor source of DHA and that, though humans are rather poor DHA synthesizers . . . the low LC(n-3)P synthesis rate may still provide us with sufcient LC(n-3)P status (p. 184). The ability of infants and adults to biosynthesize DHA makes dietary sources of LC-PUFA unnecessary for brain growth and development, even to achieve very large brain sizes.

neonates. American Journal of Clinical Nutrition 71: 1262S69S. i n n i s , s . m . 2003. Perinatal biochemistry and physiology of long-chain polyunsaturated fatty acids. Journal of Pediatrics 143(4):S1S8. kitajka, k., l. g. puskas, a. zvara, l. j. r. hackl e r , g . b a r c e l o - c o b l i j n , y. k . y e o . a n d t . f a r k a s , 2002. The role of n-3 polyunsaturated fatty acids in brain: Modulation of rat brain gene expression by dietary n-3 fatty acids. Proceedings of the National Academy of Sciences,U.S.A. 99:261924. koletzko, b., c. agostoni, s. e. carlson, t. clandinin, g. hornstra, m. neuringer, r. u a u y, y. y a m a s h i r o , a n d p . w i l l a t t s . 2001. Long chain polyunsaturated fatty acids (LC-PUFA) and perinatal development. Acta Paediatrica 90:46064. larque, e., h. demmelmair, and b. koletzko. 2002. Perinatal supply and metabolism of long-chain polyunsaturated fatty acids: Importance for the early development of the nervous system. Annals of the New York Academy of Sciences 967:299310. m a r t i n , r . d . 1983. Human brain evolution in an ecological context. New York: American Museum of Natural History. . 1995. Evolution of the brain in early hominids. Ossa 14: 4962. muskiet, f. a. j., m. r. fokkema, a. schaafsma, e . r . b o e r s m a , a n d m . a . j . c r a w f o r d . 2004. Is docosahexaenoic acid (DHA) essential? Lessons from DHA status regulation, our ancient diet, epidemiology, and randomized controlled trials. Journal of Nutrition 134:18386. p e e t , m . , 2004. Nutrition and schizophrenia: Beyond omega-3 fatty acids. Prostaglandins Leukotrienes and Essential Fatty Acids 70:41722. r o b s o n , s h a n n e n l . 2004. Breast milk, diet, and large human brains. current anthropology 45:41925 smit, e. n., i. a. martini, h. mulder, e. r. b o e r s m a , a n d f . a . j . m u s k i e t . 2002. Estimated biological variation of the mature human milk fatty acid composition. Prostaglandins Leukotrienes and Essential Fatty Acids 66:54955. smit, e. n., h. a. woltil, e. r. boersma, and f. a . j . m u s k i e t , 1999. Low erythrocyte docosahexaenoic acid in malnourished, often breast-fed, Pakistani infants: A matter of concern? European Journal of Pediatrics 158:525-26. woltil, h. a., c. m. van beusekom, a. schaafsma, f . a . j . m u s k i e t , a n d a . o k k e n . 1998. Long-chain polyunsaturated fatty acid status and early growth of low birth weight infants. European Journal of Pediatrics 157:14652.

On Hunting and Virilocality1


References Cited
b o u w s t r a , h . , d . a . d i j c k - b ro u w e r , j . a . w i l d e man, h. m. tjoonk, j. c. van der heide, e. r. boersma, f. a. j. muskiet, and m. hadders-alg r a . 2003. Long-chain polyunsaturated fatty acids have a positive effect on the quality of general movements of healthy term infants. American Journal of Clinical Nutrition 78: 31318. h a c k , m . , d . j . a n n e r y, m . s c h l u c h t e r , l . c a r t a r , e . b o r a w s k i , a n d n . k l e i n . 2002. Outcomes in young adulthood for very-low-birth-weight infants. New England Journal of Medicine 346:14957. h e i r d , w. c . 1999. Biological effects and safety issues related to long-chain polyunsaturated fatty acids in infants. Lipids 34: 20714. h o r n s t r a , g . 2000. Essential fatty acids in mothers and their

keith f. otterbein Department of Anthropology, State University of New York at Buffalo, Amherst, NY 14261, U.S.A (keitho@ buffalo.edu). 6 viii 04 Marlowes (CA 45:27784) analysis of marital residence among hunting and gathering bands fails to examine the conditions under which virilocal residence occurs. A major component of his analysis is that foragers (the term
1. I am indebted to my wife, Charlotte Swanson Otterbein, for critical commentary and editing of this report.

This content downloaded from 200.26.133.57 on Fri, 29 Nov 2013 20:20:43 PM All use subject to JSTOR Terms and Conditions

Volume 46, Number 1, February 2005 F 125

table 1 Subsistence Technology and Marital Residence


Percentage Hunting and Fishing 70100 2060 Total Marital Residence Multilocal (or Uxorilocal) 11 12 23 F p .26, x2 p 2.366, p ! .20

Virilocal 9 3 12

Total 20 15 35

he prefers) in the Standard Cross-Cultural Sample are usually multilocal (43%), infrequently virilocal (34%); uxorilocal residence is even less common (23%). His usage of virilocal encompasses patrilocal and his usage of uxorilocal encompasses matrilocal.2 He offers ve reasons that multilocal residence prevails. He offers no explanation for the fact that one-third of the foragers in his sample are virilocal. I suggest that the gure is higher. Julian Steward (1955) argued that bands were patrilineal (by which he meant patrilocal), and then his student Elman Service (1962) argued that they were patrilocal. The explanation offered was that a group of related males is the most effective hunting unit. Extrapolating back in time led to the argument that in areas where hunting, particularly the hunting of large land and marine animals, occurred, early humans were organized into patrilocal hunting bands. Service used the terms virilocal and uxorilocal (1962:30). Recently Virginia Kerns has argued that bands were not patrilineal or patrilocalthat Steward spent his career looking in vain for such bands (2003:31922)and offered reasons similar to Marlowes that patrilocal residence is infrequent. The subsistence codes that Marlowe employs demonstrate that there is great variation among foragers as to whether they primarily hunt, gather, or sh. Indeed, I think that it is misleading to call all the societies in his sample foragers, since foraging suggests gathering rather than hunting. Many of Marlowes foragers hunt. I prefer to call the peoples in his sample hunter/gatherers. Depending on the context, shing can be like hunting (if marine animals are sought) or gathering (if dead sh are collected after a pool has been poisoned). Many of the peoples in the sample that have high shing scores hunt marine animals, among them the Copper Eskimo, the Haida, and Bellacoola. I hypothesize that if gathering dominates, the society will have multilocal (or uxorilocal) bands, whereas if hunting and shing dominate, the society will have virilocal bands. To test this hypothesis, I combined hunting and shing and then dichotomized the variable so that societies with a combined hunting and shing score of 70% or above were contrasted with societies with a combined score of 60% or below. Table 1 shows a weak
2. For a discussion of the distinctions and their history see Otterbein (1978:7680).

positive relationship which reaches the .20 level of probabilityapproaching but not reaching statistical significance. Approximately half of the societies in which hunting and shing dominate are virilocal (9 out of 20). If transfer after rst years (Marlowes v216) is used as the basis for distinguishing virilocal residence from multilocal residence (the classic matrilocal-to-patrilocal form of residence which Murdock terms matri-patrilocal residence [1949:17]), 8 multilocal societies become virilocal societies. In these societies a man knows that he will be returning to the band of his birth with his wife and childrenhis loyalties lie there. Certainly he visits from time to time and discusses with male kin his future place in their band. He knows that upon his return to the band he will join them in hunting pursuits.3 Table 2 shows a strong positive relationship that is statistically signicant at the .02 level. Three-quarters of the societies in which hunting and shing dominate are virilocal (15 out of 20). Although Marlowe shows that % of contribution is not correlated with residence (p. 281), his codes show that the more hunting and shing, the more males contribute to the subsistence of bands (see also 2003:293). Available animal and plant life and subsistence technology appear to have a strong but not determining inuence upon residence. I believe that this analysis is the rst to show that hunting and virilocality are signicantly related. Further, it should be noted that a majority of foragers are virilocal; Marlowes table 2 shows 57% virilocal, undermining his basic argument.4 Certainly we should seek the causes of virilocality as well as the causes of multilocality. In summary, the percentage of hunting and gathering bands that are virilocal ranges from 34% to 57%, depending upon which residence variable is used. Arguing about whether multilocal or virilocal residence is more important in terms of frequency is a waste of time; what is important is to identify the conditions under which each type occurs. Marital residence is related to subsis3. Male cooperation is, of course, most efciently practiced among men who have grown up together in the same locality and know each others habits and capabilities and who trust each other (Service 1962:48). 4. Using the Ethnographic Atlas, Carol Ember found essentially the same percentage (1975:440).

table 2 Subsistence Technology and (Ultimate) Marital Residence


Percentage Hunting and Fishing 70100 2060 Total Marital Residence Multilocal (or Uxorilocal) 5 10 15 F p .42, x2 p 6.076, p ! .02

Virilocal 15 5 20

Total 20 15 35

This content downloaded from 200.26.133.57 on Fri, 29 Nov 2013 20:20:43 PM All use subject to JSTOR Terms and Conditions

126 F c u r r e n t a n t h ro p o l o g y

tence technology. Hunting and shing are associated with virilocal residence, and gathering is associated with multilocal and uxorilocal residence. Virilocal residence keeps a male kinship group intact; multilocal and uxorilocal residence are likely to scatter related males, preventing the development of a localized male kinship group.5

Reply
f r a n k w. m a r l o w e Department of Anthropology, Harvard University, Cambridge, MA. 02138, U.S.A. (fmarlowe@fas. harvard.edu). 5 x 04 I am pleased that my analysis of forager residence patterns (Marlowe 2004) has stimulated further analysis by Otterbein. I am perplexed, however, by his rejection of the term foragers, which I used for those societies that practice little or no animal or plant domestication (Kelly 1995). Hunter-gatherers is an equally appropriate term so long as no one thinks that it excludes shers. Otterbein says that foraging suggests gathering as opposed to hunting. If true, we need to change that impression, since optimal foraging theory applies to foraging for leaves, fruit, nectar, insects, sh, or mammals. I would agree with Otterbein that it is more important to explain the variation in residence among foragers than to characterize the modal pattern were it not for the fact that human foragers are so often said to be as male-philopatric as agricultural societies, which if true could have profound evolutionary implications (Ember 1978, Rodseth et al. 1991). One study found that three Thai hill tribes known to be matrilocal showed less mtDNA variation than Y-chromosome variation, while the opposite was true of three others known to be patrilocal (Oota et al. 2001). The signicance of this result is that such genetic analyses must be picking up the effects of recent residence patterns (unless we are prepared to think that these two agricultural hill tribes have maintained distinct ancient preagricultural patterns, which seems unlikely). The greater mtDNA than Y-chromosome variation that has been found in several agricultural societies (Seielstad, Minch, and Cavalli-Sforza 1998) therefore tells us little about preagricultural sex-biased dispersal. The importance of my analysis was foremost to counter the false impression that foragers are as virilocal as nonforagers. Otterbein says that a majority of foragers are virilocal; Marlowes table 2 shows 57% virilocal, undermining his basic argument. Apparently, he has missed my basic
5. I have argued in many places that patrilocal residence produces fraternal interest groups and that fraternal interest groups are linked to feuding and internal warfare, a relationship recognized by Marlowe (p. 281). In a forthcoming book I have argued that fraternal interest groups, weapons, and hunting form an eternal triangle that occurred in some bands in the Paleolithic (2004:62).

argument, which was that previous analyses are awed because they have used only the later years of marital residence (57% virilocal) rather than both early and later years of marriage, which yields 34% virilocal, 23% uxorilocal, and 43% multilocalmaking foragers much less virilocal than agriculturalists. Foragers look even less virilocal (14%) when we consider strict residence where no frequent alternative pattern exists. Otterbein says that I offer no explanation for why some foragers are virilocal; I do, however, offer ve reasons that agricultural societies are more virilocal than foragers, and these should be relevant to explaining virilocality among foragers. One reason is smaller home ranges and defensible resources, which should favor inheritance and male philopatry. Otterbein shows that when hunting and shing together account for 70% or more of subsistence, virilocality in later years of marriage is signicantly more common. He says that many of the foragers who have a high percentage of shing hunt large marine mammals. However, the Standard Cross-Cultural Sample codes the Haida and the Bellacoola as deriving most of their subsistence from anadromous shing, that is, catching salmon running upstream (variable 858 in the World Cultures CD [2001]). Anadromous shing is the opposite of hunting in that it leads to the smallest home ranges while hunting leads to the largest home ranges and greatest residential mobility (Kelly 1995, Marlowe n.d.). Otterbeins addition of shing to hunting implies that where male contribution to subsistence is high residence is virilocal. Too often it is assumed that hunting and shing and male contribution to subsistence are synonymous, but Australian females often hunt reptiles and Andamanese females catch seafood. In reality, where shing is more important male contribution to subsistence is signicantly higher (r p .461, p p .005, n p 36) but not where hunting is more important (r p .165, p p .336, n p 36), since in those societies female gathering is usually even higher. If Otterbeins logic is that when male subsistence activity is more important co-operation among male kin and virilocality will be favored, then surely male contribution to subsistence is a better variable to use than percentage hunting and shing. I looked in detail for effects of male contribution to subsistence on marital residence and found nothing. When male contribution is classied the way Otterbein classies hunting and shing (70%), there is no significant association between high male contribution and virilocality even in later years of marriage, the variable Otterbein still seems to prefer (x2 p 1.18, p p .278, df p 1). The same is true if 60% is used. There is, however, a signicant association between virilocality in later years of marriage and anadromous shing (table 1). Anadromous shing was the primary mode of subsistence for the Haida, Bellacoola, Twana, Yurok, Aleut, Eyak, and Klamath (v858). Most of these are considered complex foragers. Because they harvested rich supplies of salmon, which they smoked and stored, they could live in small, defended territories, in permanent villages, at high population densities with social straticationall features

This content downloaded from 200.26.133.57 on Fri, 29 Nov 2013 20:20:43 PM All use subject to JSTOR Terms and Conditions

Volume 46, Number 1, February 2005 F 127

table 1 Residence Mode After First Years of Marriage (v216) and Andromous Fishing for Subsistence (v858)
Anadromous Fishing Primary Mode of Subsistence No Yes Total Not Virilocal 14 1 15 x2 p 3.9, p p .048

World Cultures. 2001. Journal of Comparative and Cross-Cultural Research 12(1).

Virilocal 13 7 20

Total 27 8 35

On the Denition of Shamanism


pieter jolly Department of Archaeology, University of Cape Town, Private Bag, Rondebosch 7701, Cape Town, South Africa (jolly@age.uct.ac.za). 26 viii 04 Klein and Staneld-Mazzi (CA 45:4045) cite Lewis-Williams as saying that shamamism is restricted to huntergatherer societies. In fact, Lewis-Williams (2002:132) remarks that shamanism is not so restricted, as Furst (2002) has pointed out and a survey of the literature on shamanism will conrm (see Thorpe 1993, for example). Quite a number of agriculturist and pastoralist cultures have shamanic features. Since the rites and experiences associated with shamanism differ considerably, even within the classical shamanic complex of Siberia and Central Asia (Siikala 1985, 1992), close denition of the term is difcult. Nevertheless, there are some features which are common to all shamanic cultures. The following could perhaps serve as a denition: cultures that can be classied as shamanic are those which, as a minimum requirement, possess religious functionaries who draw on the powers in the natural world, including the powers of animals, and who mediate, usually in an altered state of consciousness, between the world of the living and that of the spiritsincluding the spirits of the dead. This excludes those religious functionaries, such as Christian priests, who do not draw on the powers of nature, including animals, and it incorporates the meaning of the word from which shaman is derived, the Tungusic term saman, denoting as a noun one who is excited, moved, raised and as a verb to know in an ecstatic manner (Grim 1983:15)that is, through an altered state of consciousness.

more typical of agricultural societies than of foragers (Testart 1982). While Otterbeins hypothesis about why hunting and shing favor virilocality is not borne out by male subsistence contribution, it led me to discover the anadromous-shing effect. When the anadromous-shing complex foragers are excluded, foragers are even less virilocal measured by my preferred variable (early and later years of marriage) than I had concluded. Only 8 societies (30%) are virilocal, while 7 (26%) are uxorilocal and 12 (44%) are multilocal. I thank Otterbein for stimulating this discovery.

References Cited
e m b e r , c a r o l r . 1975. Myths about hunter-gatherers. Ethnology 12:43948. k e l l y, r . l . 1995. The foraging spectrum. Washington, D.C.: Smithsonian Institution Press. k e r n s , v i r g i n i a . 2003. Scenes from the high desert: Julian Stewards life and theory. Urbana: University of Illinois Press. m a r l o w e , f r a n k w. 2003. The mating system of foragers in the Standard Cross-Cultural Sample. Cross-Cultural Research 37:282306. . 2004. Marital residence among foragers. current anthropology 45:27784. . n.d. Hunter-gatherers and human evolution. Evolutionary Anthropology. In press. m u r d o c k , g e o r g e p . 1949. Social structure. New York: Macmillan. o o t a . h . , w. s e t t h e e t h a m - i s h i d a , d . t i w a w e c h , t . i s h i d a , a n d m . s t o n e k i n g . 2001. Human mtDNA and Y-chromosome variation is correlated with matrilocal versus patrilocal residence. Nature Genetics 29:2021. o t t e r b e i n , k e i t h f . 1978. Comparative cultural analysis. New York: Holt, Rinehart and Winston. . 2004. How war began. College Station: Texas A & M University Press. r o d s e t h , l . , r . w. w r a n g h a m , a . m . h a r r i g a n , a n d b . s m u t s 1991. The human community as a primate society. current anthropology 12:22154. seielstad, m. t., e. minch, and l. l. cavallis f o r z a . 1998. Genetic evidence for a higher female migration rate in humans. Nature Genetics 20:27880. s e r v i c e , e l m a n r . 1962. Primitive social organization: An evolutionary perspective. New York: Random House. s t e w a r d , j u l i a n . 1955. Theory of culture change: The methodology of multilinear evolution. Urbana: University of Illinois Press. t e s t a r t , a . 1982. The signicance of food storage among hunter-gatherers: Residence patterns, population densities, and social inequalities. Cultural Anthropology 23:52336.

Reply
cecelia f. klein, eulogio guzma n, and maya stanfield-mazzi Department of Art History, 100 Dodd Hall, University of California, Los Angeles, CA 90095-1417, U.S.A. (cklein@humnet.ucla.edu). 16 ix 04 Jolly offers a thoughtful and in some respects fresh response to our recent reply to David Lewis-Williams, who in the June 2004 issue of current anthropology responded to our earlier (2002) article criticizing use of the term shaman in scholarly studies of Mesoamerican art.

This content downloaded from 200.26.133.57 on Fri, 29 Nov 2013 20:20:43 PM All use subject to JSTOR Terms and Conditions

128 F c u r r e n t a n t h ro p o l o g y

Like so many scholars he is still unwilling to abandon the term altogether and so proposes a new denition of shamanic cultures. His citation of Lewis-Williams to the effect that shamanism is not restricted to hunter-gatherers does not really engage and therefore does nothing to invalidate our lengthy argument that there are no reliable criteria for diagnosing shamanism cross-culturally. For Jolly, these are cultures in which religious functionaries mediate between living human beings and the spirit world by drawing on the powers in the natural world, including the powers of animals. His denition of shamanism therefore differs from more traditional denitions in emphasizing the source of the powers of shamans rather than the means by which they acquire them. It is clear that our article failed to convince Jolly that there is no consistent set of shamanic features that relate more complex societies in any meaningful way to simpler, hunter-gatherer societies. He takes such features as a given in his comments. Although he apparently does not see any danger in lumping together disparate kinds and levels of social organization under the single rubric of shamanic cultures, we continue to view this kind of homogenization as a mechanism, however unintentional, for othering the cultures included in the category. This perceived danger was a major impetus for our article that Jolly simply does not engage. Fortunately, he does not include among the most important of these shamanic features either adherence to a particular cosmological schema or the use of hallucinogens, neither of which, we have argued, can be traced across the board. Moreover, to his credit he evades the problems inherent in the notion of a mandatory altered state of consciousness by acknowledging that not all shamans profess to enter into such a state. He does, however, cling to the notion that such altered states are usual in shamanic cultures, thereby ignoring our homage (p. 388) to Roberte Hamayons point that there is no empirical way to conrm or deny the existence of a trance. We also offered (p. 388) some examples of individuals often described as shamans who either do not profess to have attained an altered state of consciousness or make no semantic distinction between waking and dreaming. Since the claim to have entered into an altered state is found in societies of all levels of complexity, we do not see it, even if only usually found among shamanic cultures, as a meaningful means of linking them together. We do nd intriguing Jollys suggestion that what binds shamanic cultures together is their belief that certain individuals can work magic by means of their special relationship with the animal world. A wide array of cultures, both past and present, seem to meet this criterion. In Mesoamerica alone, this belief has been well documented time and again. However, the Aztecssource of the slippery word nagual, which we (2002:39192) showed had nothing to do with a belief in an animal co-essencewere not among them. Moreover, Mesoamericans seem to have come up with a variety of ways of relating to animals, in some places believing that certain individuals could transform themselves into animals and in others that virtually

everyone has an animal (or plant or other entity) alter ego. Where everyone has an animal alter ego, a relationship to a special animal obviously cannot serve as a diagnostic of a shaman. Moreover, we know that priests in some highly complex, centralized societies also claim to have a special relationship with one or more animals, as was the case in ancient Egypt, where some gods took the form of an animal or a part-human, part-animal being and their priests appear in art wearing headdresses representing those animals. It seems to us that, for Jollys denition to work, the exact nature of the relationship between man and beast would have to be more carefully thought through and much more sharply dened. Most important, to our mind, is the fact that Jollys proposed denition of shamanic cultures is based solely on ideological grounds. He assumes from the start that the shaman is a religious practitioner, a bias that we criticized in our 2002 article as eliding the social and political roles played by the individuals identied as shamans. By excluding from his denition of shamanic cultures the political and economic roles of the individuals he describes as religious functionaries, Jolly implies that their control of the spirit world was simply that and not a special means of implementing and maintaining social and possibly political control. It remains our opinion that any viable denition of shamanic cultures would have to take into account how the actions of the shaman serve the interests of a particular person or group. For us, the real base of power always lies in the material world. Thus, while we greatly appreciate the effort and creative thinking that Jolly has put into coming up with a new and, in some ways, improved denition of shamanism, to our minds he has neither fully confronted nor solved the problem.

References Cited
f u r s t , p . t . 2002. Comment on: The role of shamanism in Mesoamerican art: A reassessment, by Cecelia F. Klein, Eulogio Guzman, Elisa C. Mandell, and Maya Staneld Mazzi. current anthropology 43:383419. g r i m , j . a . 1983. The shaman: Patterns of Siberian and Ojibway healing. Norman: University of Oklahoma Press. klein, cecelia f., eulogio guzman, elisa c. m a n d e l l , a n d m a y a s t a n f i e l d - m a z z i . 2002. The role of shamanism in Mesoamerican art: A reassessment. current anthropology 43:383419. k l e i n , c . f . , a n d m . s t a n f i e l d - m a z z i . 2004. On sharpness and scholarship in the debate on shamanism. current anthropology 45:4046. l e w i s - w i l l i a m s , j . d a v i d . 2002. The mind in the cave: Consciousness and the origin of art. London: Thames and Hudson. s i i k a l a , a . 1985. Comment on: Mental imagery cultivation as a cultural phenomenon: The role of visions in shamanism, by R. Noll. current anthropology 26:44361. . 1992. Finnish rock art, in Studies on shamanism. Edited by A. Siikala and M. Hoppa l, pp. 5667. Helsinki: Finnish Anthropological Society/Budapest: Akade miai Kiado . t h o r p e , s . a . 1993. Shamans, medicine men, and traditional healers: A comparative study of shamanism in Siberian Asia, southern Africa, and North America. Pretoria: University of South Africa Press.

This content downloaded from 200.26.133.57 on Fri, 29 Nov 2013 20:20:43 PM All use subject to JSTOR Terms and Conditions

You might also like