DRUMS AND GUNS: A CROSS-CULTURAL STUDY OF THE NATURE OF WAR by Valerie Wheeler Nammour A DISSERTATION Presented to the Department of Anthropology and the Graduate School of the University of Oregon in partial fulfillment of the requirements for the degree of Doctor of Philosophy December 1974 ii APPROVED: iii Copyrighted by Val.erie Wheeler Nammour November 8, 1974 iv VITA ... NAME OF AUTHOR : Valerie Wheeler Nammour PLACE OF BIRTH: Norwalk , Ohio DATE OF BIRTH : July 19, 1940 UNDERGRADUATE AND GRADUATE SCHOOLS ATTENDED : Oberlin College Unive rsity of Nebraska University of Oregon DEGREES AWARDED : Bachelor of Arts With Dist i nction , 1962 , University of Nebraska Master of Arts , 1967, Uni versity of Nebraska AREAS OF SPECIAL INTEREST : Cross-Cultural Methodology Economic Anthropology Indians of North America Phi l osophy of Social Science Violence and Armed Conflict Women ' s Studies PROFESSIONAL EXPERIENCE : Graduate Assistant, Department of Anthropology , University of Nebraska , Lincoln , 1962-1963 Assistant Curator of Anthropology , Uni versity of Nebraska State Museum, Lincoln , 1963-1 966 Instruct or , Lane Community College , Eugene, Oregon , 1966-1967 V Graduat e Assi stant , Department of Anthropology , University of Oregon , Eugene , 1968-1 969 Lecturer , Departm~pt of Anthropology , Sacrament o State College , Ca lifo~nia , 1970-1 972 Assi stant Pr ofes s or of Ant nropology , Californ ia State Unive r s i ty , Sacramento , 1972- AWARDS AND HONORS : Sigma Xi vi ACKNOWLEDGMENTS Thi s endeavor has taken seve r al years and great a mounts of energy , others ' as well as mine . I am indebted t o several people, all of whom deserve specific mention : t o Professor Joseph G. Jorgensen , who suggested a large cross- cultural study of warfare as my dissertat i on topic; to · Professor Ri chard P . Chaney, who helped me design the research and guided me through its pi tfall s to i ts conclu- s i on ; t o Professor Vernon R. Dorjahn , for shepherding me t hr ough the intricacies of my relat i onships with the National Science Foundation and t he Graduate School , and for his cr i t - ical attent i on anu gentle pressure ; and to Professor Malcolm McFee, for his enthusiasm and constructive criti cism. Above all, I thank my ent i re committe e for allowing me the intel- lectual freedom to go beyond a strict quantitative study and expl ore quest i ons of how we know as well as what we know . I coul d not and woul d not have finished the work wi th- out t he support of my family . Richard Page Whee ler and Lucy Pope Whee l er , my parents , have set examples and given unflag- gi ng encouragement . My husband, Jamil Nammour, and my son Edouard have live d and breathed the entire effort i n good spirits , and Professor Nammour nas proved to be my most capabl e research assistant . Finally , I wish to thank t he National Science Founda- tion for its research support in Grant GS- 3234, and Caro lyn C. Larsen for her masterful production of the final copy of the dissertation . vii -TABLE OF CONTENTS Page LIST OF TABLES . ix LIST OF FIGURES xii LIST OF MAPS . . . . x i ii I NTRODUCTION 1 Chapter I . AN ESSAY ON WAR 6 Is This Really Science? .. 7 The Problem of Definitions · 17 Why Do ~eople Do It? ... . 28 Are wa rs natural? ..... . 29 It is i n our minds ...• 47 Frustration l eads to aggression . 47 People learn how to be aggressive . 53 Cultural determinism and war 62 Etics , emi cs, and laws . . . 64 War and the exercise of power . 1 31 Answers . . . . . 1 57 II . METHODOLOGY 182 The Logi c of Cross-Cultural Studies • · 182 What are we looking for? .... . . • 182 How is it done? .... . 189 Who has found what? .... . . . Coding and Stat istical Pr~cedures .. The data . . . . . . . . . . . . ··.: mg Statistical ruanipulation of t he data • • • 251 The Variables ..... . ... . . . • . 254 I I I . FINDINGS : DISCOVERIES OF THE ODYSSEY 279 A View of the World . . . . . . . . 279 Africa 280 Circum- Mediterranean 282 Eastern Eurasia . 283 Insular Pacific 284 North America . 285 viii Chapter Page South America .... 286 Wor l d Confi gurations 288 Regional Confi gurations . 369 Africa ...... . 371 Circum- Medi terra nean 385 East Eurasia 393 Insular Pa cific . 400 North America 408 South Ameri ca . 418 FINAL REMARKS 429 NOTES 432 APPENDICES 437 A. Standard Cross-Cultural Sample . . . . 438 B. Code Sheet for ODYSSEY .... 451 C. Coding of Va riable s for the Standard Sa mple 461 Comparison of Phi Coeff icients Among Geographica l Regions ... • .... 481 E . Bibliographic Sources Use d i n Coding Warfare Variables Listed by Standard Sample Identity Number of Ea ch Society . . . • • 503 REFERENCES CITED ......... . 533 iX ·•. LIST OF TABLES Table Page 1. Who Calls What Violence 58 2 . Divale's Sex Ratio Data 118 3 . Relation Between Continents and Warl ikene ss 203 4 . Comparison of Q Coeffibi ents : Geographic Region as a Variable .... 281 5 . C Coefficients for the World : Each Variable with Every Other Variable .. 289 6 . Intercorrelation of Polygyny and I nternal War 296 7. I ntercorrelation of Polygyny and Externa l War-Attacking .............•. 299 8 . Intercorrelations of Twelve Other Variables with Polygyny .... . ....... . 300 9 . Intercorrelation of Settlement Pattern and Subsistence ... . ....... . 311 1 0. Intercorrelat ion of Settlement Pattern and Mean Community Size . ..• 312 11 . Intercorrelat ion of Subsistence and External War-Attacking . . . . . . . . . . . . .. 313 1 2 . Significance Levels of Associations i n the State Configuration . . . . . . . . . . . . 316 13 . Intercorrelation of Levels of Juri sdictional Hierarchy Beyond the Local Community and External War- Attacking ....... . 318 14 . Intercorrelation of Levels of Jurisdictional Hierarchy Beyond the Local Community and External War-Being Attacked ........ . 320 X Table Page 15 . Intercorrelation of Levels of Jurisdictional Hierarchy Beyond, the Local Community and External War~Attacking ....... . 322 16 . Intercorrelat ion of Levels of Jurisdictional Hierarchy Beyond the Local Community and External War- Being Attacked ...... . 323 17 . Signif icant Correlations with O Lambdas .. 327 18 . Significant Correlations with Lambdas in Two Directions ........ . . . . . 333 19 . Intercorrelat ion of Military Organization and Political Organizat ion . .' . . . . . 335 20 . I ntercorrelation of Military Expectations and Political Organizat ion ...... . 336 21 . Intercorrelation of Military Success and Political Organization ........ . . . . . 337 22 . Intercorrelations of Eight Other Variables with Mode of Succession of Local Headman . . . 340 23 . Intercorrelation of Two Classes of Military Expectations and External War . . . . . . . . 347 24 . Intercorrelations of Specific Military Expec- tations and External War-Attacki ng . . . . . 348 Intercorrelat ion of External War-Attacking and Externa l War-Being Attacked ..... . 350 26 . Reconstruction of Otterbein ' s Intercorre lation of External War-Attacking and External War- Being Attacked . . . . . . . . . . . 353 27 . Intercorrelation of Agriculture and the Military Expectation of Subjugation 361 28 . Intercorrelation of Intensive Agriculture and the Military Expectation of Subjugation . 364 29a . Intercorrelat ion of Population Density and Internal War . . . . . . . . . . . . 368 xi Table Page 29b . I ntercorre l at i on of Population Dens i ty and Externa l War7Attacking ..... . 368 30 . Excerpts of Regional Comparisons of Phi from Appendi x D ........... . 370 31 . · Comparison of Driver and Schuessler ' s I nter- correlations (Pearson ' s r) of t he Si x Geographica l Areas wi th Thos e Obtained from the Standard Sample ........... . 372 32 . Signifi cant Corre l ations wi th O Lambdas-- Africa • . . . . . . . . . 380 33 . Significant Corre lations with Lambdas in Two Directions . . . . . . . . . . . 381 34 . Africa : Intercorrelation of Levels of J uris- dictional Hierarchy Beyond the Loca l Community and External War-Attacking . . . . . . . . . . . 384 35 . Circum- Medite1ranean: I ntercorrelat ion of Levels of Jurisdictional Hierarchy Beyond the Loca l Community and Exte r nal War- Attacking . . . 391 36 . Significant Corre l a tions with Lambdas in Two Di rections--Cir cum- Mediterranean . . . . 392 37 . Signifi cant Corre l a tions with Lambdas in Two Di rections--East Eurasia . . . . . . . . . . . . 399 38 . I nsular Pa cific : I ntercorre lat ion of I ntensive Agr iculture and External War- At tacking 406 39 . Si gni f icant Correlations wit h Lambdas i n Two Directions--Insular Pacific .. .... . 407 40 . North America : I nt ercorrelation of Levels of J urisdictional Hi er archy Beyond t he Local Community and External War - Attacking 41. Significant Corre l ations wi th Lambdas in Two Di rections--North America ......... . 417 42 . South America : I ntercorrelation of Military Succecs and External War- Attacking ..... 424 4 ;7;. . Significant Correlations wi th Lambdas i n Two Directions- - South America ...... .. . . 426 xii LI ST OF FIGURES Figure Page ·, 1 . The World : Direct ion and Strength of Lambdas , Focused on 2~4 Leve l s of Polity . . . . . . . . 326 2 . The World : Di rection and Strength of Lambdas , Focused on 1-4 Levels of Polity . . . 331 3 . Africa : Direction a nd Strength of Lambdas , Focused on 2-4 Levels of Polity. . . . . . 375 4 . Africa : Direction and Strength of Lambdas , Focused on 1-4 Leve l s of Polity . . . . . . 378 5. Circum-Mediterranean : Direction and Strength of Lambdas , Focused on 2- 4 Levels of Polity . . 387 6 . Circwn- Medi terranean : Direction and St rength of Lambdas , Focused on 1- 4 Levels of Polity . . 389 7 . East Eurasia : Direction and Strength of Lamb- das, Focused on 2-4 Levels of Polity . . . . • 396 8 . East Eurasia : Direction and Strength of Lamb- das , Focused on 1-4 Levels of Polity 398 9 . I nsular Pacific : Di re ct i on and St rength of Lambdas , Focused on 2- 4 Levels of Polity 403 10 . Insul ar Pacific : Dire ction and Strength of Lambdas , Focused on 1-4 Levels of Polity . . . 404 1 1 • Nort h America : Direction and Strength of Lambdas , Focused on 2-4 Levels of Polity 4 11 12 . North Ameri ca : Direction and Strength of Lambdas , Focused on 1-4 Leve l s of Polity 413 13 . South America : Di rection and Strength of Lambdas , Focused on 2- 4 Levels of Polity 421 1 4 . South America : Di re c tion and Strength of Lambdas , Focuse d on 1-4 Levels of Poli ty . . . 423 xiii LIST OF MAPS 1 • Africa . . . . . . . . . 374 2. Ci r cum- Mediterranean . . . 386 3 . East Eurasia . . . . . 395 4. Insular Pacific . . . 402 5. North America . . . . 409 6 . South America . . . . . . 420 INTRODUCTION Throughout 1he history of t he discipline , anthr opolo- gi sts have shown a perenni a l interest i n warfare , generally avoiding the quest i on of uni versal causes of war and ins tead des cribi ng specific practices i n specific societies . These descriptions vary i n thoroughness , refle ct i ng both the emphas is upon warfare and war - related. a ctivit ies of i ndi- vidual cultures , and the e ra and i nterests of the ethnog- raphers . The quali ty available to the cross- cultural re search worker ranges from the full-blown accounts of Nor t h American I ndi ans in t he Great Pla ins during the 19th century to mere passing comments fo r many s ociet ie s in Asia and Africa . Between 1950 and 1960 , only a few anthropological papers on warfare and one ethnological study, Turney- High rs Primit ive War (1 949 ), were publi shed . Duri ng the de cade 1960 to 1970 , t he literature grew, largely stimul ated by concern over the I ndochina war . In response to a demand wit hi n t he anthropological community, a symposium on war was held during the American Anthropological Association 1 s national meet i ngs i n the f all of 1966 . Subsequent ly , the papers and some comments from the f loor were published (Fried , Harri s , and Murphy 1 968 ) . It was f e ~Lt at t he time , and by a reviewer l ater , t :Lat t he results of the symposium were "not what we have been wai ting for " (Fox 1969 : 315) . 2 Since 1966 the national meetings have each i nc lude d symposia on warfare and conflict , yet i nspe ction of paper abstracts indicates that the emp~asis has continued to be on the spe- cific case and the ·spec' ific society , with a reluctance to generali ze . Those generalizations that have been made are sweeping ones , chara cterized by the judgment that human a ct i ons are explained best by determini st and unitary theori es . After a l engthy discussion of what i s right and wrong about these contemporary generalizations , along with the problems inherent in generalizing in social science , I have presented yet another attempt at generalization , usi ng the rationale and methodology of cross-cul tural re search . The issue of why me~ kill each other is too complex to be ana- l yzed solely i n terms of a few we l l - chosen cases , either single cases or a small sample subjected to statistical anal- ys i s . Large amounts of data can be handled through stati sti- cal analysis , the method used i n this dissertation . The body of literature i s small but growing , and the styles in quanti- fication have undergone changes . Quincy Wright (1942 ; rev . ed . 1965) provided tabulations early on relationships between "warlikeness " and continental location , temperature , natural habitat , climatic energy , race, subsistence , political organ- ization , social organizat ion, and cultural isolation for 650 societies . He made little attempt to define these variables , and he di d not perform any statist ical manipulations . Broch 3 and Galtu.ng (1966) recently used. Wright rs data to measure associat i ons be tween cultural complexity and frequency df warfare , us i ng an ui nd'ex of pri mitivity . 11 They found that t he "level of belligerence varied from O per cent at the most primitive level to 95 per cent at the least primi t ive or tra- ditional level. " All other recent re search has been carried out usi ng smal l samples : Naroll (1 966 ) made a cr os s-cultural survey of 48 primitive societies on t he frequency of warfare and mili t ary orientation i n order to te st t he deterrence hypothesi s : that societies with strong mi litary orientations (armaments , fortification , tactics) will engage less fre- quent l y i n war . He found the oppo s ite to be the case . A subsequent test ( Naroll 1969 ) of 30 hypotheses rela ted to deterrence using a sample of 20 hist or ic civilizat ions pro- duced s i milar re sult s . Otterbe i n and Otterbe i n ( 1965 ) pub- lishe d a cross- cultura l study of 50 societies on the rela- tionship of feuding t~ the freque ncy of warfare , fraternal interes t groups , and l evel of political complexity . The ir results s howed that feu ding occurs when frate rnal interest groups , indicated by the presence of polygyny and patrilocal- ity , are present but that i t is controlled by politica l authority and declared s tates of war . Otterbein ( 1968 ) also trie d to demonstrate , again using 50 socie ties , that frater- nal groups and unauthorized raiding parties i nf l uence t he frequency of i nternal war i n uncentralized. pol.i ti cal systems but not i n centralized one s , and that the frequency of 4 external war doe s not i nf lue nce t he frequency of i nternal war . Ot terbe i n in a l a rger wo rk ( 1970), but s t i l l using a s ample of 50 societies·~ tested t he sur vi va l va lue of milita ry ' efficiency i n the mai ntenance a nd evolut ion of po l iti ca l com- munities . Ember and Ember (1 97 1 : 593 ), us ing Otterbein ' s war- fare codings and with s ampl e size vary i ng from 18 to 33 societies , found that 11matri loca l soci eties have purely external warfare [ that i s , warfare only with other societies ] much more often t han patriloca l societies ,tt and "if warfare i s continual, men will contribute more than women to subsi s - tence unless the warfare prevents t hem from doing so ." I will discuss these cross-cultural studies , particularly, Otterbein ' s work , in great er det ail in Chapter II . My own preliminary work on pri mitive warfare i ncludes (1) a small study ( 35 soci eties ) whose central variable i s the part icipation of women in warfare , testing 7 hypotheses based upon vari ab l es of military expectations, frequency of war , leve l of political complexity, dominant subsistence a ctivity , dominant di vision of l abor , residence , and des cent. ( 2) I have carri ed out indi re ct stat i s tical te s ts of my s ample and Otterbein ' s 1965 and 1968 sample s fo llowing the rationale of Cha ney and Revilla (1 969 ), a nd I have found sim- i l arly tha t stati s tica l generalization base d on s mall samples will diffe r cons i derably from tho s e based on large ones (in t he te s t ca se , 412 societie s ). The c r ucial iss ue of s ampling and sample s i ze i s a lso di s cuss ed i n det a i l i n Cha pter II . 5 I n this dissertation I have carried out a l arge- s ca le i nductive i nvest igation of variabl es that may be associated wi th warfare among l argely preindustrial peoples . Rather than be concerned with the problem of cause , for reasons pre- sented in Chapter I, my contribut ion i s to make s ome general- i zations about the strengths of association between warfare and othe r sociocultural phenomena , based on a l arge standard sampl e of 186 s·ocieties evenly di stri buted over the 6 ge o- graphic r egions of t he worl d (Murdo ck and White 1969). The use of such a l a r ge s tra tified sample allows one ~o explore geographic vari at ions i n the patterning of vari ables . There fore , t he problem under i nvestigat ion is t wofold : how does one ask t he quest i on "Why war? " anthropologically a nd what ki nds of answers does one get using cross-cultural quantification? I n worki ng toward solutions to the problem, my purpose is to make warfare i ntelligible and t o a ccount satisfactor ily for human a c t ion . 6 CHAPTER I AN ESSAY ON WAR 11 You shoul d l earn not to make personal remarks ," Al ice sai d with s ome s everi ty : "it ' s ve ry rude . 11 The Hatte r opene d hi s eyes very wide on heari ng this ; but all he said was , 11 Why i s a raven l ike a writing- desk? 11 "Come , we shall have some fun now!" thought Alice . 11 Jt m glad they tve begun asking riddles-- I beli eve I can guess that , 11 she added a loud . 11 Do you mean that you t hi nk you can find out t he answer to it ? 11 said the March Hare . "Exactly so ," said Alice . --Alice in Wo nderl and The ri ddle is war . What is it? Who does i t? When do they do i t? How do they do i t? But the most important ques - tion , Why do they do it? , will have to go wobbling around on weak knee s fo r reasons I hope to make clear i n this chapter . It means tha t I shall offe r no macro- gene ralization , no l aw- l i ke statements , no cause- a nd- effect statements , no functional statements , and no predictive models . We l l , why not? Aren t t I supposed to be s cient i f ic? That ki nd of contribution to knowledge am I maki ng , anyhow? How can my work be useful towards eliminating warfare as the s courge of mankind? I ndeed , others have tackled t he why quest i on , and their argu- ment s will be pre sented below and anal yzed fo r what appear to be their strengths and weaknesses . But underlyi ng all attempts to tackle , anal yze , and argue is , i n my estima tion , 7 the a cceptance or reje ct i on of determinism in social science . I t i s best to begin at the beginning . . Is ~his Really Science? To be sc ientifi c means trying to f i gure out the world around us or , more weighti l y , explaining the empirical rationally. Later on I shall dis cuss a useful cr iterion to distinguish between s cience and pseudo-science, but what I set forth here i s the basic philosophica l i ssue of this essay . The empiri ca l refers t o experience i n t he _world--few woul d dispute t hi s --but rationality i s obvious l y le ss clear . We a ll claim to use it, yet we come up with competing and conflicting expl anations of what appears to be the same experience . Irra tionality is ignorance , i nsanity ... or re ligion and art . We do not expect t he artist to be rational, at least not in the same way as the philosopher or scient i st , and he is the antithesi s of the scientist . In our Western t radition of science and philosophy, we believe that the exercise of reason upon the worl d will reveal order, necessi ty , cause-and-effect, and predictability . Anything less , we think , l eaves us with randomness, accident , unpre- dictability, and chaos . As arguments , the former is called determinism and the latter, indeterminism. The force of my criticism of various theories of war i s directed aga i nst determinism in a variety of f orms , and towards a third view that I believe i s the reasonabl e way to make human behavi or 8 intelligible , the primary business of anthropology . Those who deal with the concept culture inevitably must confront the issue of individual free will and responsibility , although many ma.y avoid coming to grips with it. Since inde- termi nism i s a negat ive theory and is parasitic on determin- ism (Ayers 1968 : 6) , as relativism is parasitic on absolutism, one cannot understand it without knowi ng what determinism-- the most fundamental and i ntimate issue i n human behavior--is all about . We ordinari ly take it for granted that most adults are normal and that a l l normal adults are respons i bl e for their actions . Yet there are powerful arguments to sh9w i) that responsibility is incompatible with determinism , since if determinism is true , no one can act different ly from the way he does ; but ii) that responsibility is also incompat ible with indeterminism , since if human a ctions are not caused at all , not even by the agent, then no one is r esponsible for them ; and iii) that there i s no third alternative (Ba ier 1970 :1 00 ; italics added) . Or , in greater detail , what Ayers calls the Basic ArguJnent for determinism reasons thus . The common premiss of both determinist and indetermini.st i s , of c ourse , the alleged incompatibility of causation and freedom of choice. This supposit i on i s usually sup- ported by the argument that , if an event is in all respects causally explicable, then it could not have been any different, since to explain an event completely just i s to shew that nothing eise was possible in the circum- stances . If one thing is the whole c ause of another , then given the occurrence of the first thing , the other must occur . This is the meaning of 1 cause '. Likewise , to say that somethi ng is a law is to say that anything different is i mposs ible . So i f everything that happens is governed by a law , nothing that fails to happen is possible , or ever was possible . On the other hand , every- one agrees , or should agree , that for there to be freedom of choice or, therefore , any real choice at a ll, the agent mus t be presented with alternative s that are all genuine possibilities . It follows that a free choice 9 ca nnot be causally determined , and it ca n have no c omple t e explanation (Ayers 1968 : 1- 2) . But inde termj_ni sm cannot account for free choice , either , and the de termi ni st se'ems to hol d the high card by cla i mi ng that the logica l consequences of indetermini sm l ead to absurdity , re futing his opponent thus : Let us suppo se that some human actions are not caused , that they are inde termined and so really could have been otherwi se . This i s simply to suppose them accidental , random , unpredictable and unre l ated to t he agent t s per- sonali4; y . How can we even ascribe such an event to an agent as an action of his , unless we can relate i t to a specif~c and i nte lligible motive , and how can we do this if i t is a bolt from the blue? ..• Why should a man be held responsible for something supposed to be unre lated to all antecedents and perhaps to all that comes after- wards, s omethin~ that came into his head from nowhere? (Ayers 1968 : 3-4) . Most of the bas ic tenets of the concept culture are ant ithet- ica l to even the possibility of such bolts from the blue : behavior is shared i n common with and lear ne d fro m other human beings. Ye t anthropologists s till assume that indivi d- ua l s are re spons ibl e and do have freedom of choice wi thin the context of a spe cific culture . The only conclus ion left to draw , i t is argued , i s ..• tha t responsibility and freedom of choice are se l f - contra1ictory not ions , requiring that an agent both could and coul d not have a cted otherwise , that hi s action both was and.was not causally determined . Determini sm . .. is no longer prese nted as the l e s ser of two evils , the more acceptable horn of a di l emma , but as t he only pos s ible c oncl usion of a rigorous argument (Ayers 1968 : 4) . Indetermin:ilsm is di scarded as irra tional and determinism ret a ine d as r a t i onal , albeit nonhuman . This vari at ion of i nhumanity seems to be mor e tolerable to phil os ophers and 10 scientists t han the imputed chaos of i ndeterminism . 1 Ayers (1 968 : 4 ) makes a distinction between metaphysi- cal determinism and s cientific determinism : the former is " the do ctrine that no person i s ever responsible for his actions or ever makes s free choice "; the l atter i s " the v iew that every event has a sci entific explanation ," i . e . , it is "explicable by reference to i ts antece dents and l aws of nature .... 11 That i s all he has to say on the mat ter of s ci- entifi c determinism, since his task is confront i ng the me ta- phys ical. He doe s not specify whi ch sorts of scientifi c subje ct s he is i ncluding--physical, social, or both--but jf one makes reference to l aws of nature , he must have a l ready de cided that such a c oncept i s appropri a te to the problem. at hand , i . e . , he has a lready a ccepted metaphysical de terminism . Since Ayers ' l arger t ask i s to de f end i nde termi nism yet make it responsible , he i s tal king about people . Ob jects are not r esponsible . Therefo re , while not applicable to the physica l sciences , his conce rn with the nature of human a cti on is . obviously relevant to the social sciences . Even if it were limited to phys i ca l s cience , his st atement obscures the dif- ficuJ_ties of "cause " and "laws of nature " in talki ng about science younger than Newtonian me chanics . For some anthropologists con cerned with t he ory, the propriety of either scientific or metaphys ica l de t erminism i s not at a ll clear . For others where theory , there determi ni sm . For example : 11 Every ethnographer--including , of course , t he humaI'-ist - literary- natural history-" slice - of-life " kinei--makes some attempt to tell us how the fact s hang together , hm-T they i nterrelate causally, and what makes the part icul ar socie ty i n question what i t i s . Yet to assert, for exampl e , t hat a particul ar event , x , occurred because of certa i n other events , x and~, pre- supposes , more broadly , that events of type~ are linked to events of type y and ~ i n cert a i n determinate ways (Manners and Kaplan 1968 : 5 ) . Not only do anthropologi sts not speci fy the dete r mi nat e the- ories i mplicit in the i r ethnographies , what little anthropo- l ogical theory that does exist i s flawed , i n Manners ' and Kaplan ' s assessment, by sub jectivity, ignorance , and i deology . They i ns i st that, i n order to be s cient ific, anthropologi sts must c onform to the hypothetico-deductive reasoning of formal logi c. That i s , facts and empirical gen- e ralizations can only be considered to be explained when they are subsumed unde r t heoret ical statements , either deductive ly or probabilis tically , and t hus their occurrence made predict- able . They dismi ss as subje ct i ve , and therefore i mmature , speaki ng of theories as maki ng something "intelligible " or "understandable ," and evaluating theori es on the basi s of which " sati sfie s " us t he most (Manners and Kapl an 1968 : 7 ), enamoured as they are wi th posi t i vist notions about how we should know the worl d , rather than how we do know t he worl d ,. Nor do they c onsider the differences between socia l and phys i cal theory to be i nherent , but due i nstead to the i nadequacies of t he i nves tigators ; 12 It is not our i ntent ion .•. to suggest that such differ- ences [between the phys ical and social sciences] in any way constitute a logical or ontological gap between the two fields of i nquiry . Paradoxically enough , greater theoretical demand~ are made on the social s cience s than are made on the physical sciences .... Thus , one looks to the social sciences for a nswers to the many social problems that afflict us , and one wants detai le d answers so that sornething may be done to correct them (Manners and Kaplan 1968 :1 0) . To which one can only r eply, the social need for predictive power does not mean it is possible . The only way I can see that social theory can be predictive i s to force people to behave according to the theory . Marvin Harris , as the clearest proponent i n anthropol- ogy of the new s cientism, cultural materialism , and cultural determinism, is devoted to scientific determinism--which he formulates only as similar variables under similar condi tions tending to give rise to similar consequences- --and he , like Manners and Kaplan , c onsequently must view human behavior i n terms of metaphysical determinism . Harris acknowl edges the classi c implicat ions of determinism: If i ndividual behavior is l arge l y a predictabl e outcome of technoeconomic, technoenvironmental , and other given conditions, what significance does anthropology attribute to the strivings of i ndividuals to change their person- alities or to modify their cultures? Are we all autom- atons fated to act out our particular predestined personal and culture configuration? Is our sense of free will merely an illusion? Can we hold ourselves and each other responsible for the choice of personal and cultural life- styl e that we exhibit? (1971 : 593) Yet he di smisses them with the cla i m thcq.t" t he determin- i sm governing sociocultural phenomena is a matter of proba- bility rather than of certai nty ," and of low probability 1 3 leve l s at t hat , "a lthough our batt i ng average woul d be qui te re spectable among meteorologi sts and ge o l ogi sts . . . . 11 By extens ion then , predicting human behavi or i s i n t he same class as predicting the we athe r and earthquakes . He attributes exceptions : ... errors may be made i n data c ollection and processing ; the statement of the initia l co nditions may be inad- equate ; the conditions may be undergoing evolutionary change ; and fi nally, the generalization may be poorly constructed . All of these s ources of error may be r educed to one : lack of sufficient i nf ormation , or incomplete knowle dge (1 971 : 594 ). Harris does not pursue further the problem of incomplete knowledge, and he concludes by reconci ling free will and determi n i sm to his satisfaction : While cultural anthropology is predicated upon the gen- eral s u bordination of t he individual to the force s of encu ltura tion, every s ociety nonetheless exhibit s a wide s pectrum of i ndividual personalit i es . Enculturat ion i s a form of programming , but our knowledge of the . content of the program i s a l ways quite incomplete . Thus no i ndivi d- u a l , even t he most heavily psychoanalyzed , possesses any- thing more than a probabili s tic k nowledge of how he will a ct under g iven contingencies . But at the same time , it i s perniciousl y fal se to sup~ose t hat a ll sociocul t ural events are equally p robable and that by mere force of wi l l the i nsp i red i ndividual can alter the tra jectory of an entire sociocultura l syst em in a d i rection c onveni ent to any philosophy . Conve r gent and pa r a llel traje c tori es far outnumber divergent traject - ories i n soci o cultural evolution . Most people are con- formists . Hi story repeats i tself i n count less acts of indi v i dual obedi ence to cultural rule and pattern, and indi vidual wills seldom prevai l i n mat ters requiring radica l alterat ions of de eply c onditioned beliefs and practice s ( 1971 : 595- 96 ) . Probably most anthropologists woul d agree with the general drift of Harris 1 first paragraph , a lthough he assumes the u nknowab l e --that cu ltural programming i s c omple t e , although 14 we can never know its full content . But i n the se cond para- graph he raises the specter of chaos and god-like omnipotence and then counters it by stat i ng that indi vidu_al free will cannot override the forces of history, i.e., the wills of other people , something tha t no indeterminist woul d cla im. Furthermore, he assumes that parallelism and convergence of sociocultural trajectories i s due to l aws of nature , not the spreading of i deas over time and space . The gist of all this i s that Harris is a determinist, but be cause he i s also a humani s t, he i s a probabilistic determini st who does not offer a level of ac ceptance or rejection of a generalization. Do 2 or more events have to appear together in a cluster (independent of other clusters, of course ) most of the time , some of the time, 2 time s out of 3, 50 per cent of the time , 30 per cent, 10 per cent, or 5 per cent before one can claim the di scovery of a causal chain with an arrow poi nting i n one direction? On the one hand , those whom Ha rris calls historical particularists--Boas and his students primari ly--would never claim that culture mani - fests itself randomly or deny that things go together i n pat- terns that may crop up in disparate parts of the world . On the other hand , the logica l positivists would simply reject Harris' use of predictability as unscientific . Bidney (1 967) offers one explicat ion for our feelings of agreement and disagreement with Harris ' conc lusions above : 1 5 Paradoxical as i t may appear , cultural de terminism and cultural i ndeterminism are not i ncompat i ble i n practi c e , since the concepts refer to complementary f a ctors i n the l ife of man and soc i ety . Cul tural i ndeterminism is pos- tulated be cause culture i s not a closed system , but r a t her an open one , subjec t to the directive agency of human i nte l l i gence . On the other hand , some degree of cu l tural determi ni sm characterizes human l ife , ana the cu l tural ant hropologi sts may study comparatively the s i g- ni f icant correlat ions between culture , personality , and s oc i e ty ( 1967 :1 7-1 8 ). But Bidney ' s pos i t i on i s mere l y an i nver s i on of Harri s ': Harri s a llows t he i ndivi dual some freed om wi thi n t he deter- mi ned processes of cu l ture ; Bi dney denies t hat de termini sm but a lso deni es i ndi vi dual freedom . Fi nally , he does not g o or~ to expl ain what woul d c onst i tute a "signi f i cant corre- l ation ." Harri s (1 968 : 300 ) a br upt ly di smi sses Bi dney ' s ve r - sion of cultural indetermi nism a nd label s Bi dney rs cl a i m that culture is the product of human freedom i n creat i v i ty and choice as " the reductio ad absurdum of cul tural indetermi n- i sm ." In Harri s t metaphysics cul ture and freedom are contra- di.ctory and yet , as we have seen , Harri s does admit t hat ?,t l east some of t he time human be i ngs are fre e. Berreman , in a bri ef brilliant art icle (1 972 : 224 ) s ees the situat i on as a dilemma i~ terms of science and humanism : " ... how to be s c i entif i c and at the same time retai n t he humanist ic i nsights--the human re l evance - - wi t hout which no account of human beings makes sense ," a dilemma that anthro- pologists solve by avoidance , that i s , choos ing e i ther s ci- enti sm or humanism . The former he i dentifi es as 1 6 ... a retreat to , or preoccupation with, such thi ngs as quantification, abstract models , simulation, and highly formal met ho ds of data collection and analysis . It reject s intuitive insights ... [and] results most often in descriptions and interpretat ions which are reliab l e but whose validity is que st ionable .. •• The l a t t er, humani sm, he accuses of reject i ng ... any serious attempt at scient i f i c method, relyi ng ent i re l y upon intuitive i nsights and the qualitat i ve , . empathetic ethnographic result thereof .•. [ and ] results i n ac counts which may be valid , but whose reliabili ty i s undemonstrated . To reso l ve the dilemma , Berreman proposes that anthropolo- gists deve l op a method that 11 c ombi nes rigor and i ns i ght, verifi cation and di scovery , accuracy and empathy , replica- bi l ity , and human relevance ," and he ca lls for a "so ciology of ethnographi c knowledge " or an "ethnography of an ethnog- raphy ," i n whi ch the ethnographer speci f i es the bases of hi s i ntuitive and t heoret i cal inferences , hi s procedures , and h i s sources (1 972 : 228 ). He does not , however , i ndi cate how detai l ed the ethnography of the ethnography must be i n order for veri f i cation t o set i n . He a ccuses ethnoscience of pro- liferat i ng means of explanat ion into such torturous channel s that explanat io n i s eit her never achieved or is i nhuman, a view with which I concur . He l i kens the difference i n points of view between scientist and humanist t o the difference between a European navigator of the open seas , whd charts a line to his goal "according to certain uni versal principles " and subse quent l y bends his efforts to "remaining on course " or foll owing a pl an and secondarily to reaching a distan t 17 landfall, and a Trukese navigator , who knows where he is going and gets there by continually taking i nt o a ccount the context of his actions--t he prevailing condit ions of wi nd , waves, tides, and currents . The basic principles of the Europeans are easily taught, but one must be an apprentice to learn Trukese ways (Be rreman 1972 : 225), that is, one mus t learn the contexts . Yet European and Trukese eventually arrive at other places and they c an tell you how they did i t , but which one will tell you more a bout the world in which he lives? Both are science . This i s s cience , too . Wel1 then , what is this phenomenon--War? The Problem of Definitions "Define your terms " i .s the usual admonition at the outset of a scientific investigation . It is assume d that one cannot or at least should not proceed unt il this is taken care of. Since argument commonly arises over matters of def- inition--often to a stalemate-- we use op~rational defi nit ions , specifically designed for the case at hand and perhaps con- taining "objective " measures of some sort . I n submitt ing to this traditional demand , I advocate a definition of 1·mr that has 3 necessary conditions: (1) people organized i nto a polit i cal group , although "political" seems redundant ; (2) . an i ntention or expectation of the group to do harm to the body or belongings of another group, for which (3) armed 18 conflict i s the means . Succinctly , war is collective arme d assault . The i nter.t ion , especially if people go to war only for revenge and defense , may be simi lar to that of a feud . I · have i ncluded r a iding , which may be cons i dered as violent a ction midway between feuding and warri ng , as a form of war- f are. My j ust ificat ion for this defini t ion is that violen t behavior i s a co ntinuum and that it is difficult to spe cify qualitat ive differences between one form and another. I think tha t the fundament a l characte r of warfare i s violent action t aken by one group of people agai nst another , however they defi ne themse lve s a s members of that group . Further- more, this intent ion is to "further the i nteres t s of one group at the expense of the other t hrough willful destruction of life and goods . War expre sses the apotheos i s of se l f - i nt erest11 (Hoebel 1958 : 508 ). From s ociety to society , war to war, the s ize of the groups may vary , as may the principles of organization that unite the members , the weapons and tac- tics of armed confrontation , and the self-i nterest s a t issue, which c an range from defense i n the face of overt attack to outright c onquest of l and , expropri at ion of go ods , and enslavement of peoples . So far , thi s definit ion seems commonsensi cal, even obvious . Should i t be more specific? Should I stipulate that a f i ght can only be called a wa r if there is a mini mum of n peopl e i nvo l ved ; or t hat the result is a mi nimum of n casualties (lethal or non-lethal); or spe cify t he type , 19 val.ue , or quantity of what is gaine d ; or require that certain principles of organization be present ; or that a certain di s - tance be maintained between combatants using certain weapons ; or that host il.itie s be concluded according to certain pri n- ciples--do we recognize a war by how it turns out? Several anthropologists have tried to establish criteria to identify 11 true" war . Turney-High (1949 : 21-22) li s ts 4 conditions of war~ a group motive rather than merely an individua l one ; leadership providing command and dire ction; tactical oper- ations designed to bring warriors to advantageous positions ; and the abil.ity to sus tain an assault until the goal of the war is a ttained . Bohannon (1963:306) identifies war by its ends : "True warfare," if we may call it such, ... has as its ends peaceful settlement with new political conditions, not continuation of fighting .... when we find it, we are going to find societies in which it i s carried out by speci alist bodies ca lled "armies " and others i n which it i s carried out by whole bodies of citizens. Newcomb (1950), in a typology of warfare , says that true war exists only where people are food-producers and engage in ..• a type o~ armed conflict that t akes place between soci eties , meeting in compet ition for anything that is valued by the groups involved, usually consisting of territory or certain products of this territory, such as good hunting grounds , oil-producing or agricultural lands (p. 317) . According to his typo:iogy based on technology (Newcomb 1960) , foo d- collectors (gatherers, hunters , parttime horticultural- ists ) do not engage in 11 t rue 11 warfare . The logical i mplica- tions of this 1ine of reasoning mean that food-col1ectors , 20 e . g ., most of the Indians of North America , carried out non- war , or "false " war , or that whe n the Hopi fought t he Navaho , the Hopi were at war but the Navaho were not , or that the U. S . Army was carrying out true war agai ns t the Dakota , Cheyenne , and Apa che , but the I ndians were not . This appli- cation i s not meant to be facetious , but to illustrate the limitations of what appear to be carefully stipulated defi- nitions when they are he l d up to the real world . By hi s definition true war exclude s food-collectors and beginq with the agricultural r evolution. Yet .Newcomb a lso i ncludes hunting grounds as something of val ue over which people fight . Therefore we shoul d be able to i nclude, for instance , horse nomads of the Great Pla ins. But Newcomb has previously put such hunters i nto his Type 2 or "primi- t ive " warfare category , below the category of true warfare carried out by food- producers . The Patwin, Wintun, and Miwok of the Central Valley of California fought and killed gro\lpS of intruders in their gathering g rounds . They and othe r California tribes carried out ext ensive r evenge raids and even pitched battles . But people at this level of techn ology and subsi stence would belong in Newcomb ' s fir st and lowest category , thos e pe oples i dentif ie d as not havi ng warfare . Wha t then do we do with such violence? These examples should show some of the limitations of this particular typology . We know that there exist differ- ence s of soce sort i n the warfare practices of these 21 societies , and it may be the case that economi c driteria are the most fruitful for des i gning a typology. As McEwen (1 963 : 160) says, one is forced to make a choice as to what element is to be the most important and which ones must remain sec- onda ry . At the very least , a typology would give us linguis- ti c means to talk about variations i n warfare in a more abbre- vi ated fashion . Yet I do cons i der t a lk about "true " warfare , like "true " reli gion , to be l udicrous . Furthermore , it i s often hard to push cases i nto the given s lots . I do not f i nd that the vi rtues of Newcomb ' s typology outweigh its l imi ta- t i ons and sugge s t that we look elsewhere . Not onl y must we worry through definit i ons about whe ther or not a state of war exi sts in order to begin count - i ng cases, but also whether or not the study of war in pre - industri a l societies i s relevant to understand i ng ( 1) war carried out by nation- states , regardless of the time peri od , and (2) war carried out by nation- states in the modern age . These two questions have been asked simultaneous l y ( see Gorer below ). The implications of the first question are that war s waged by Egypt of the Old Ki ngdom , Rome of t he Caesars, Engl and of Henry II , Zululand of Chaka , the American Civil War , and the recent engagement of the United State s in I nd o- china a ll belong in the same category , and t tat this category is quali tatively different from one i n which we might put the Aztecs (but not the Inca), the Comanche , t he Rwal a Be douin, and the Dagum Dani . Fo r many critics , the wars of the l atter 22 societie s are hard to i dent i fy and different from our wars . For most of humani ty , the tribe i s the unit within whi ch killing is considered murder , and outside which killing may be proof of manhood and bravery , a pleasure and a duty . Such ki.lling may be done by i ndi viduals--head- hunters , scalp- col lectors , as part of a vendet t a or rai d--or by groups ; in the l a tter case the killing i s ca l led "warfare ." The differenc EB in quality and s c ope between tri bal warfare and modern war between nat i on- states are so great that it might be useful i f different words were used for the two a ctivi ties (Gorer 1966 : 31 ) . Added to the diffi cul ty of i dentifyi ng pri mi tive pe opl es i n general and i n part icul ar i s the di ffi cul t y of i denti- fy i ng t he i r wars . Primitive peopl es only rare ly conduct f ormal hosti l it i es wi t h the ob j ect of achi evi ng a t an- gi bl e economic or poli ti ca l result . The i r hos t ili t ie s are se l dom conducted by a highl y organized profess ional military class us i ng dist i nctive i nstruments and tech- n i ques regul ated by an i ntergroup law applicab l e only duri ng the period of "war" and designed to r ender war an eff i ci ent instrument of poli cy . The se e l ement s which go t o make up the concept of war today are product s of civ- i l i zat ion , and onl y the i r rudi ments can be fo und among pri mitive pe opl es (Wright 1965 : 58 , italics added ) . As for Wri ght ' s criterion of forma l host ilit i es , I must ask what would an informal war be l ike ? What ki nds of differe nces are we dealing with i f one distinguishes primi t i ve conflict as ritualized and civili zed conflict as f ormalized? Two of t he forms t hat are i mportant to peopl e i nvo l ved i n contem- porary wars are declarations of war and t reat i es of peace . We are a l l fami liar with the semantic howlers that Korea and Viet Nam were not wars because Congress did not de clare t hem to be so . (Cf . Donald Wells , The Wa r Myth .) I must also ask , when does "modern" begin? Hi s t or i ans identify "modern history " as the most recent phase of a period that began with the Renai ssance and Reformation and 23 ended wit h the onset of "contemporary history, " which "begi ns when the problems which are actual in the worl d today f i rst take visible shape 11 (Barraclough 1967 : 20) . Barraclough him- se l f argues cogently that the years be f ore and after 1890 mark the division between modern and contemporary h i story , although the latter term may be a contradic tion , and we may use "era" instead (1 967 : 20- 21 ). Furthermore , what do we do about Medieval wars? The Crusades of the 11 th century , the Holy Wars of Chri stendom , are the direct ancestor of total war , a unique practice of the Western world . After 105 3 , the idea of Holy War made rapid progress . Precedents there may have been ,, but in it s formulation and execution it was something entirely new . Hitherto , i n common with the other great religi ons , Christianity had condemned war as essent i ally evil , and t he Eastern Church continued to maintain it s reservations . War might be unavoidable ; it could not be good , still l ess coul d it be holy . How, then , are we to explain the ease wi th which the West cast aside so hall owed a tradition? (Barraclough 1970 : 13) Barraclough goes on to suggest that contact between Chri st ian and Norse pagan transformed each : the Norse warriors accepted Christ , but the Chri st ians accepted Thor and Odin . Gentle J esus became a warrior chief dedicated to conquering, not converting, the rest of the worl d (1970 : 13-14). After 1053 Holy War became the battle cry of the Papacy , an i nstrument for extending its power and authority ... [ and ] it had nothing to do , i n i nception, with the s t rug- gle against Islam .... We no longer regard the Crusades ... as a great movement in defense of Western Chri stendom , but r ather as the man- ife station of a new , driving , aggressive spirit which now 24 became the mark of Western civilization . We no longer regard the Latin states of Asia Mi nor as outposts of civilizat ion i n a worl d of unbelievers , but rather as radica lly uns table . centers of colonial exploitation ..•• There is perhaps a , line running from BoI'-emund of Ant i och to Cecil Rhodes (Ba rraclough 1970 : 14 ,1 6) . Benedict (1959) uses a simpler taxonomy than t hat of Newcomb to distingui s h primitive from modern war , 20th century war in this instance . She follows the gene ra- species model of natura l h i story . I n her view , the characterist ic of the genus War "is homicide that is rewarded with unquest ioned acclaim and gratitude by one r s fe llows ". ( 1 959 : 370-71 ) , as cont rasted to the genus Murder , which i s homicide with the heaviest penalties . At the species level, primitive war1are is "socially nonlethal, 11 that i s , combat ants exist i n self- sufficient societies and their wars "do not drag to ruin the civilization of both tri bes that engage i n them" (1 959 : 374 ). Modern war , however, i s socially l ethal to all parties, s i nce it shreds t he fabric of i nterdependence t hat exists among modern nations and thus i s suicidal-- You cannot be an international civiliza tion and reap its benefits , and at t he same t i me engage as a nat iona l to destroy other nationals root and branch . War in such a society becomes a case of cut ting off your no s e to spite yo~r face (1 959 : 378- 79 ) · --and drags your civilization down in ruins . Nonetheless , the views of Gorer and Wright seem r epre- sentative of anthropologica l thinking , a vi .ew set probabl y by Malinowski (1 941), who def i ned wars in authentically pri mitive societies as highly ritua lized , self-limiting , and 25 qualitatively di stinct from wars i n our time . As other i nstances , Vayda (1 960:1 - 2 ) and Chagnon (1968a ) i dent ify primitive warfare through the fo llowing criteria : ... s mallness of scal e in military operations , short duration of active hos tilities, poor development of com- mand and discipline, great reliance on stealth a nd sur- prise, and the great significance of village community or loca l group in organi z i ng and conducting war parties (Chagnon 1968a :110). Another often used critical point of difference is distance between combatants , i.e., primitive war is pe rsona l because i ndi v i dual s at least see each other and may even have physica l contact; modern war i s i mpers ona l, because combat - ants drop bombs on mere topography and shoot artillery from ranges of several mile s . Diamond (1 968 ) makes it the differ- ence between personal and i mpersonal evil. Yet what do we do a bout labeling the war i n Viet Nam , where t he acts agai nst the enemy that have raised t he greatest outcry at home have been acts of personal evil? It is reve aling that we judge such acts as "murder" and therefore outside the category of war , which l eaves i ntact the myth of i mpe rsonal combat. We need not go to the more dramatic forms of personal contact. I nfantrymen do have at l east visual contact with others l abe led as "enemy" and they do l egitimately kill. I s this not personal? There are really two different types of human extermi- na tion i nvolved in Vietnam , and t hey perhaps require two different kinds of expl anation . Firs t, there is extermi- nation such as t he Huey troops engage in--extermination at clo se range , i n which the killer can see (and enjoy , apparently ) the blood he sheds . Second , and f a r more 26 c ommon , there is extermination at a di stance , i n which the extent of t he killing is so vas t t hat t he killer tends to think i n teri1tS of areas on a map rather t nan i ndi vi dua ls . I n neither case i s the victim perceived as a person ( such a perception woul d make modern war i mpos- s i ble) , but in .the first case the killer at least sees the imme diate consequences of his act , whereas i n the second case he does not . ·:rhe "close- range II killers in Vietnam are confronting something . ... (Slater 1970 : 41 ) . Applying Slater , is Vietnam s chizophrenic t hen--two wars in one? Or should we deal i n percentages , i . e ., impersofial v i o- lence i s more common ( ove r 51% of the time ), and file this war i n the modern ca tegory? I am attempting to show through simple examples that the problems of ''when is a war a war and when is i t not a war? " and "when is a war primitive and when is i t modern? " are not to be solved through sett ing forth defi nitions . I contend that Anthony Leeds ' protes t in the 1967 AAA Sym- posium--that 11 we have had no conside ration of what the phenomenon war is and what it is not . What s ort of i nstitu- tional complex is it generically? 11 - - is a que st i on that we real ly do not have to answer before getting on with the work . Actua l l y we a lready have an answer . When an ethnographer describe s warfare in a non-We stern society , he has neces- sarily recogni zed and understood an event as being war because it ties into his prev ious experience (cf . Winch 1958 ). But a r ej oinder might be that understandi ng i s rela- tive and a ll definitions are equally valid ... and u l timately that cross - cultural compari sons are i mpossible . What one 27 ethnographer recognizes as war , another does no t. Do we dwi ndle into cultural solipsism? But how do we recognize war in the first place? We do it not on the basis of the presence or absence of 1 or 2 variables , but on how those variables look against a back- ground or in a context of human activity and ideas, somet i mes difficult to i sol ate . We know what "war " is--we use the word competently every day and other peopl e li stening to us kn.ow what we mean ; they must , because we are still talking. "," ~ 'r j Furthermore, we recognize the metaphorical use , e.g ., the war on poverty, the war on crime , the war on inflation, i mmedi- ate ly because of the context of a decade of official concern and strategems for collectiv e assault against social enemies . But what about the phrase , t he war on health? It does not make sense , at first, until we examine the cont ext--in an article in the New York Review of Books--and find that it refers to and condemns efforts of medical entrepreneurs t o defeat measures which would ostensibly gi ve better health care to the population , regardless of financia l status. Competent usage does n ot require identical understand- ing of what const itute s war or warlike behavior a cross cul- tures . Recall i n history the chronic compla int that the eriemy does not f i ght fair, or as gentlemen , or according to the Geneva Conventi on . Indeed, identica l understandi ng is s i mply not possibl e : we cannot take a list of defi nitiona l c r iteria, and hold them up against t he world as though they 28 were color chips looking for a match- up . What would we see i f we r an 3 fi l ms--Dead Birds , newsre els from Vietnam and a pro f essional footb~ll game--fir s t without sound track , a nd a second time with? What would we see a s the s ame and a s dif- ferent? What woul~ happen if we cha nged 1 vari able , a nd sub- stituted for our Western observer a Tasaday tribesman from t he Philippines? Since the Tasaday are not acquainted with war , is he an "unbiased " observer , objective , purely empiri- cal , unpolluted? Ca n he i dent i fy the facts of the matter and tell us how the people i n the 3 "films are behaving the same or differently , without a ctua lly knowing what they are doing? Such an experiment i s within the range of possibility, and I contend that our unbi ased Tasaday would not be able to iso- l a te s i gnificant difference s without absorbi ng our culture, which of course eliminates unbiased reporting . I di scuss further definitional problems of war from t he political poi nt of view i n Section 4. Why Do People Do It? People have been providi ng causal explanations for war t hroughout hi s tory, with a variety of suggest ions . A profile of the cause s looks like t hi s : people go to war be cause of contiguity , habitua tion , soci a l l ea rning , predation , psycho- logical defense s (ra tionalization , blami :ng , de nial , counter- phobic tendencies) , t he "host of fears a s sociated with the human condition ," territoriality , power , fru stra tion , 29 biologi cally rooted aggression , i nstinct , and sadism (Gilula and Dani els 1969 :403) . I n what follows , I have tri ed to sort out these "causes 11 into l arger types of explana tion , · even though i n many i nstances the categories are not mutua lly exc l usive . I n t he intricate l att ice whi ch one i nevitably encounters i n talking about war , one can get hopelessly bogged down without some ca l mly applied heurist ic s , even a t the risk of doing some vio l ence to the material . The f ollow- ing discussion of human aggression and war will deal with (1) the biologi cal-inst i nctua l theory ; (2) psychologi9al t heorie s--frustrat ion-aggressi on , social l earni ng , a nd a daptation-copi ng; ( 3 ) ecologica l and economic theorie s ; and (4) polit ical theory . The r eader will note that the discus- sion is a r ranged in a hierarchy from i ndividual human biology to the poli tical behavior of human groups. · The discussions are criti ca l, a l though not exhaust i vely so. My task is to elucidate the st r engths and weaknesses of current theory and to i ndicate the direct i on of a possi ble integrative or alternative point of view . 1 . Are wars natural ? "Let the jury consi der their verdict, 11 the Ki ng sai d , f or about the twentieth ti~e that day. 11 No , no !" said the Queen . "Sentence first --verdi ct afterwards ." 11 Stuf'f and nonsense !" sai d Alice loudly. "The i dea of having the sentence first !" "Hold your t ongue !" said the Queen , turning purple . 30 "I won ' t ! " said Alice . 11 0ff with her head ! 11 the Queen shouted a t the top of her voice. - -Alice i n Wonderland To better understand why things happen the way they do, we are often urged first to be come fami l i ar wi th the na t u re of t he world and the events around us, i n the best manner of scientific method u s i ng reason and obse rvation . I n the interest of changing our world , presumably for t he bette-r, it may be argued that if we f i nd that i t i s not in the nature of someone or something t o behave i n a desired way , then all our best efforts wi l l be for nothing . We c om- monl y hear as a final reas on , justifi cation , and a u thori- tat i ve statement the declaration that s omething is " contrary t o one ' s nature ." "You can ' t fight nature ." "That ' s human nature for you ." When i nquiring a bout the nature of a thing or event , we are asking for it s essential characteristics and dist i nguishing qua l itie s , i t s e ssence , i ts inherent c harac- ter . As a re sult of our inquiry , one of t he essential cha r- a cteristics of t he subject under inve st i gation coul d turn out to be that it i s, or is not, natural, i . e ., inborn , dete r - mi ned , normal to the spe cie s . We can ask , what i s t he nature of love , of s teel , of organic gardening , of war . As part of our i nqui ry here , we are a lso asking , is war natural? We can lo ok for regula rities i n the vari at i ons of the three nece ssar y conditions posited fo r war- - a group wi th a common 31 intent ion and arms--and say something about the i ntricate ~eb of rela-'--ionships among technology and economics , politics , belie fs and values , and war . Or , possibly, we can even say something about the evolution of war in human history. But why the regularities in the first place ? Are they due to history and the spread of i deas f rom society to society? Are they functions of society itself? Or are they due to biologi cal evolution? Sooner or l ater in our i nquiry i nto the nature of war , we must tackle the question , are wars natural? Is violent group conflict inna te , inborn , and nor- mal to the species Homo sapiens? Basically , there are 2 schools of thought on this matter , with battle l ine s drawn , cannons at the ready. One school argues for the pri macy of biology, particularly of i ns tinct, in explaini ng human aggression , i . e ., man is by i nstinct an aggressive creature and this accounts for indi- vidual and group aggression and violence . Its major speakers are Nicolas Tinbergen (1 968) , Konrad Lorenz (1966 ), Robert Ardrey (1 96 1; 1966 ; 1970) ~ Desmond Morris (1 967 ; 196 9) ; Lionel Tiger ( 1969), and Derek Freeman (1964) . The other s chool , mobilized i n print by Ashley Montagu (1968 ) with such allie s as Geoffrey Gorer, J . P . Scott , E . Leach , Ralph Holloway , and J . H. Crook , argues for the primacy of culture or learning in explaining the rise and expression of destruct ive aggression . Montagu himself cont i nues to c on- sider it is i nte llectual and moral respons ibility to counter 32 the i nf luence of the biologica l de terminists, i n print and i n person . Let me set the arguments out in some detail . I am devot ing speci a l at tent i on to Ardrey and Lore nz because the others have spun off from them , and more importantly , be cause t hey are being referred to as expert witnesses by historians and others who understandably cannot themse l ves go thr ough . and anal yze e thological data and thus requi re some summary statements provided by specialists . Konrad Lorenz has been call ed the founder of the modern science of ethol ogy . Most of his re search has been done with tamed ducks and greyl ag ge ese . I n his book On Aggression , he trie s t o make a case of ins tinctual aggression , stating that it is spontaneous , cumu- l at ive , and can be directed towards good or bad ends . He c ont i nua l ly sees in human behavior analogies to pair- bonding , dominance , and fighting behavior of geese--and a particular simil arity between the greylag triumph ceremony and hun1an t "mi l itant enthusiasm" or call to battle , i dentifyi ng both a s i nstinctual . Lorenz does not consider militant enthusiasm an evil , but a re sponse spr inging from the same source as love and l aughter and effort expended to protect the group . But one of t he greatest human probl ems i s t hat we have not devel- oped i nstinctual ritualizations of i ntra- specific aggression . His book has been widely read , and his revi ewers pra i se his warmth and love of animals , if not his ant hropology . The greatest di ffi culty with Lorenz 1 work lie s i n 33 anthropomorphi sm : he interprets goose behavior i n human terms and then offers goose behavior as evidence that human be i ngs behave like _geese and consequent ly are subje ct to genetic programmi ng . Robert Ardrey i s a sli ght ly different case . A pl ay- wri ght who became i nterested i n the evolution of human behavior several years ago , he has produced 3 best - selling "personal inve s tigations 11 : Afr ican Ge nes i s (1961), The Ter- ritorial I mperat i ve ( 1 966 ) , and The Social Contract ( 1970 ). My part icular concern here i s with The Territoria l I mperat i ve , a c ollect ion of hypotheses which one finds difficult to tie together tha t has r e ceived an enormous pre ss . Ardrey's , pos ition i s that we can have a better understandi ng of mo dern man a nd develop more rationa l so l utions to problems if we study man t s i nst i nctual e ndowment (Gorer 1968 : 76 ), whi ch we can learn about by studyi ng the f ossil remains of pre - and ea r ly hominids and the behavior of other vertebrat e s--fr om fis h to birds to ungulates to other primates . His attitude i s that the news may be gl oomy , but we must face up to the fact that hwnan beings are subje ct primarily to biol ogical l aws . Ardrey goes beyond Lorenz and attributes human aggres - sion t o the "instinct for t erri toriali ty ," i . e . , that each human (male ) has an i nst i nct i ve dri ve to acquire , maintain , and defend territory ; that man builds a fence around his home and feels patriotism f ·or hi s country because of the same motivat i ons as those of his dog barking behind that fen ce ; 34 that f i ghting over the possession of territory i s a pm·rerful, i nevitable , ineradicable , and therefore uncontrollable i nst i nct . He flatly states , furthermore , that he sees no difference in degree between the psychological attachment of a lungfish to a stretch of muddy water and t he psychological at tachment of t he resident of San Francisco to the city he loves (1 966 : 337 ) . Ardrey cannot find evidence among other primates , especially the great apes , to support his thesis , but he accounts for this gap by declaring man ' s nearest phy- logenet ic relatives to be "evolutionary fai lures ." Ardrey further postulates 3 basic needs in H. sapiens, listed i n order of importance . First , the need for i dent ity and assertion of i ndividuality ; second , the need for stimu- lation and relief from bore dom ; third , the need for security . He states that 2 "institutions ," warfare and territoriality , satisfy these needs better than any others . Territoriality provides one with an area of l and (or a ir or water) with which to identify; boundaries for which one competes a nd which one explores and defends relieve boredom; one t s own place gives security . Security is the least important need becaus e in the long run security is anonymous and bori ng and will be sacrificed to experience identity and stimulation . War sat i sfies the need for identity through the possi- bility of glory and , primarily , position in rank . I n con- tras t , civilian life provj_des only unbearable anonymity . War is stimulating , the "ultimate release from the boredom of 35 normal existence " (1 966 : 336 ), and people like war be cause pain is more st i mulating t han pleasur e . Finally 1 war increases security _t hrough the acquis j_t ion of l and and loot, and t he possibility of los i ng a war is more anxiety- provok i ng than war i tself . As moder n war has gai ned i n s i ze and sophistication 1 i dentifi cation and st i mulation hav e i ncreased concomitant ly . The Pax atomica i s frustrating because ·we no l onger can have t he warfare we want . Theref ore , we must have substitute satisfactions for t hese needs , such as athletic . cont ests . War has another function i n Ardreyrs view, which is really a restatement of Thomas Hobbes ' s ocial contract and William Graham Sumne r ' s hypothesis of social cohesion (1911 ). Withi n a social group , Ardrey sees a condition of mutual hos- tility (enmity) to be the normal one , a condition re duced onl y as external hazards and t hreats to the group rise , forcing individuals to cooperate (amity ). Once external hazards are reduce d or eliminated 1 the na tura l condition of enmity will reassert itself . He even sees hostility betwe en mother and child to be inevitable once dependency is out - grown . But we still are not clear about human agg~·e ss ion agai nst fellow members of the spe cies . If everyone stayed home , assumi ng of course that male s are i n fact territorial , there woul d be no conflic t . But humans do not stay home ; they i nt rude beyond t he ir borders i nto one another t s 36 territory . Why? Ardrey traces the origi ns of murder , of man t he killer , to man ' s phyl ogenet ic history as a predator-- hunting other animals for food . I ndeed , he makes man the predator and man the killer equivalent terms . So ... the members of this school see aggression ( and war ) agai nst other humans as natural--inborn , innate , i nstinctual. Montagu and the other members of the opposing s chool i n Man and Aggression (Montagu 1968 ), mostly professional biol ogical and social s cientists , assert the _primacy of learned behavior in human beings , no t just i n recent times , but for the l ast 2-4 million years . They argue that during human evolution , learning has so suppressed any i nstinctual endowment that man has l ost virtually all i nst i ncts . The only remnants are poss i bly react i ons to sudden loud noises and to sudden withdrawal of support (Mor.:.tagu 1968). Ardrey i s accused of reviving several 19th century social do ctri~es , including Adam Smith t s I nstinct of Property, Herbert Spencer ' s Social Darwini sm , Freud ' s Death I nstinct , the Victorian thesis of the I nnate Depravi ty of Man and the Myth of the Wil d Beast , the American frontier t s Rugged I ndi vidual- i sm (including I nstinct for Status ), and t he earlier vi ew of Thomas Hobbes that the state of nature among men is a state of war , which people manage to suppress i n t he i ndivi dual by forming political groups (Montagu 1965 , 1968 ) . Ardrey rs opponents state that we have not i nherited aggress ion and 37 territoria lity from our r emote biologica l ancestors . Instead , each generat ion lea rns them from the previous gener- ation (Boulding 1968 : 88 ) . Among Ardrey ' s multitude of errors , Crook (1 968) po i nts out that Ardrey uses his basi c concept , territory, without specifying its mul tiple meanings used by ethologists . Birds , f or i nstance , have 4 types of territory : the nest or breeding area, which i s nearly universal ; t he area around t he nest which may be defended against intrus i on from male co- specifics ; an exclusi ve hunting area tha t predators oft en have and whi ch soc i a l mammals share ; a copul a ting ground or arena where only those mal es that have a cquired it are se lected for mates by the females - - thi s i s a rare form, e . g ., the Uganda Kob (an ungulate ) and New Guinea Bowerbirds . Even where these specific te rritories may be occupied and defended , such a ction takes place primarily during mating season ; for the rest of the year the animal s gather i n l arger groups ( Carrighar 1968 ). An objection to thi s could be made that si.nce human beings possess year- round sexuality , their "mat i ng season " l asts from puberty to climacteric . While one could use the terms "proprie t or" and "property " i n such con- text , as Ardrey continually does , it i s confusing because not even j_n a metaphori cal sense are these breeding or hunt i ng grounds like private property among human beings , i. e. , transferable, movable , heri table . 38 As for man the predator creating man the ki l ler , i n conter:1porary hunting societies throughout t he world fo r which ~e have informat i on , people hunt for food . The y are what I l i ke to call "prayerful predators . " They attri bute t houghts , emot i ons , and beli efs to other animals ; they thank the ani- mal s for allowing themselves to be killed ; tney treat game wi th r espe ct and ritual ; they do not t ake pleasure i n the a ct of kill i ng , nor do they hunt for fun . The l as t characteri s - t i cs se em to be those of amateur hunters , men i n f ood- producing societies who consider hunting not a necessi ty but a recreat i on . Furthermore , hunting peopl es know the di ffe r - ence between kill i ng other animal s and kill ing other human be i ngs , even though t hey may not ac cord the enemy qui te t he human stat us they give t hemselves : "I t ' s wrong to ki l l a human being but all right to ki l l an enemy . 11 How peopl e make this distinct ion between us and t hem , human being and enemy , is a tantalizing probl em whi ch I s hal l i nt r oduce at t his po i nt and return t o i n s ocial and politica l expl anations dis cussed below . Psychiatrist Erik Erikson has coined t he phrase 11 pseudo- speciation" to explai n thi s dis- tinction : members of the i n- group ident ify non- members as animals and therefore as game to hunt down and ki l l (Tiger 1969 : 213) . But thi s explanation is not adequate ; in fact , I think it in error . When men go to war , t liey know full well tha t the creatures they may kill are not bison, deer , 39 antelope , or bear . I f you ask the bi got why he hates and he re plie s , "be cause those people live like ani mals ," he knows ~ull well they do not live like the animals i n the San Diego Zoo, or in Serengeti National Park , or the hogs i n Farmer Jones ' sty , or my dog J ason . The Cheyenne called themselves the Human Beings , but surely a Cheyenne recognized a Crow or a U. S. soldier as humans too, although not as human as a Cheyenne. The American soldier may call the Vietnamese a "gook " or "dink " and hate him with fury, even decl are that he i s not human and t nerefore deserves to die , yet , upon ques - tioning , admit t hat i n fact he is a human being . While i t i s not at a ll clear how people perceive fellow humans , it seems safe to say that they do not confuse t hem with animal s hunted for food . Predat ion is a neutral term , des cribing fooa-getting activity ir.. terms of ecology and biology . "Killer , 11 however , is a moral term , and Ardrey and others who use the two synon- ymously create dramatic confusion. We speak of lions and eagles as predators and hunters , describing how they keep themse lves a live . But any time we use the word "killer ," we are maki ng a moral judgment about the right and wrong of a part icular a ct of killing o I t i s analogous to the use of the words "murder " and "stealing ." For instance , it i s incorrect to say , as many cultural relativists do , that i n some societies i t is all right to corrunit murder , or that in some societies it is all r i ght to steal . Murder and steali ng are 40 everywhere and always wrong , a l though the particul ar context i n whi ch the act may occur is variable from one culture to another (Cook n . d . ) . I f we corifuse terms and call the Cheyenne hunt i ng bison a "killer , 11 we are i mplicitly stat i ng that what he i s do i ng is wrong . If one i s a Buddhist or Hindu , this is what one means . But if we are not Buddhists or Hindus , and we are concerned wi th des cribing the probable evolut ion of the Australopithecines as savannah- dwelli ng ca r ni vore s , we do not use moral terms . In fact, use of moral terms i mmedi ate ly reveals our moral position on a particular issue . For example, in the quotation cited earlier, Slat er uses the word "killer " repeatedly : I n neither case [close- range vs . killi ng from a distance] is t he victim perceived as a person ... but in the first case the killer at l east sees the immediate consequences of his act .... The "c l ose- range " killers i n Vietnam are confronting something .. . . (1970 : 41 ) , to emphasize the immorality of the entire business . The i ssue of the confusion of terms is a crucia l one , especially i n dealing with such an emotion- charged subject as warfare . I t i s a confus i on of what philosopher Harry Nielsen calls "Socratic truths " and "s cientific truths. 11 A scient i f ic t ruth belongs to a type that one man or a few can discove r and pass on to others . The truth goes fort h " to whom it may concern 11 from di scoverers who have no way of knowing the identity of those , if any , who will be concerned . As a bit of objective truth, i t carries no stamp of its f i nder ' s i denti t y , since it i s not ab out him as an indi- vidual, although out of generos ity the world may affi x his name t o it . It says nothi ng personal to or about the individual who come s upon i t afterward , since it i s no t 41 about him e i ther , even when it i s a truth about hi s spe - cies . Finally , i t s primary i mportance is to stand avai l - able , whe ther many or few or- none ever look i t up ( rielsen 1967 : 52) . Thus , it is a scient i f i c truth that the ancestors of Homo sauiens evol ved as omni vores , dependi ng upon both vegetabl e and animal foods obtai ned through gathering and hunting . It i s a l s o a s cientifi c truth that · this subsistence pattern changed onl y 11, 000 yea rs ago , with the i nvention of plant dome s t ication . Strictly speaki ng , Homo sapiens was and i s a predator at least part of the time . But the methods of natural s cience do no t work i n t he realm of Socrat ic truths , which are of two types ( Ni elsen 1967 : 52- 53 ) . The f i rst i s beliefs and val ues that we hold personally and express in our attitudes and behavi or toward our se l ves a nd others . I t i s a truth of s ocial s cience , for i ns tance , t ha t American bl a cks have not held the same civil r i ghts as Ameri can whites . I t is a t ruth about myse l f that I do or do not care about the s ituat i on , and that t ruth ho l ds regardl ess of anyone e l se ' s opi nion ( l ike s cient i f ic truths ), I but someone e l se cannot di scover thi s truth and pass i t on to me , and since it i s a truth ab ou t me , it i s meani ngl ess sepa- r ate d from me . The second type of Socratic trut h i s a matter of faith--beliefs about ult i mate origins and v a lues , and the future that come from authorities ou t s i de the indi v i dual ,, e . g . , Jesus or Buddha . . . or Robert Ardrey . 42 The mischief occurs when someone tri es to transform a Sacra ic truth into a scientific one , or a scientific truth into a Socratic one . For example , it appears that Professor William Shockley of Stanford University be lieves that Ameri - can blacks should not have the same civil and s ocial rights and benefits as American whites because they are biologicall y i ncapable of equivalent intellectual a chievement . He tries to transform a moral truth about himself into a scientific one and thus put it beyond the realm of moral challenge . Ardrey performs a similar piece of legerdemain : he trie s to remove acts of human vio l ence agai nst other human beings from the moral realm into the scientific one , and he accuses any- one ·who challenges his claim of being a poor s cientist or a liar involved i n a co nspiracy against Truth! Transformation also can move i n the other direction . Since modern astronomy has disc overed that our galaxy i s only one of millions and our sun i s only one of trillions in a universe without appar- ent end, s ome might claim that an individual human life therefore i s insignificant and any belief to the contrary is pathetic (Nielsen 1967 : 55 ) . A classic confusion of scien- tific and Socratic truths is the perennial confrontation of Darwin 1 s theory of evolution by some bearing the Book of Genesis . Creationists are unable to see that there is no contradiction , that evolution i s scientific , and Creation is Socratic as a question of faith and not a 11 t11eory " that the Almignty offers to the public . Possibly some Creationists 43 are cauo-ht up in the mystique of science i n Western culture and cannot be content with faith alone but must find 11 proof n to buttress their faith . I have encountered religious stu- dents who refuse to talk about faith altogether ; they argue that their explanation of the world is f a ctual . As a s cientific truth, killing fellow human beings may be due to the fact that , unl ike other animals , we have not evolved inhibitions to avoid violent conflict (Gorer 1968). I n a fight between two wolves , if one wolf bellies up and bares his throat , the other one is instinctual ly inhibited from striking a death blow and the fight i s over . I t is assumed that , if wolves did otherwise , they woul d soon be extinct . Humans have a variety of submissive and surrender gestures , but unlike the wolf they may not inhibit further aggression . Why should t his be so? A possible , and untest - able , explanation i s that the human capacity to create and manipulate symbol s through language has kept the absence 9f such instincts from being l ethal to the species . Tha t is , talking to avoid or concl ude fight i ng has been successful often enough to keep ritualization or extinction from occur- ring through natural selection . Selection has given a s light edge to talking , yet not enough to eliminate lethal fights . There i s a problem with mutually unintelligible languages , yet how often do tctal strangers , without any previous knowl- edge of each other , go to war? This argument is plausible , especially when based on the premiss that language is the 44 human quality and that it has probabl y developed , as Hollo,vay contends (1969 ) , in direct association with too l - making . 2 Whi le Ardrey contends that mutual hostility i s the normal human condition, Montagu argues that coopera tion and love are the normal condition , that violence and destruction are due to l earned responses , frus t ration , and faulty val ues i n a compet itive , overcrowded , and dangerous world . War i s of re cent invention i n t he whole of human history : 12, 000 years ago i n t he Neolithic , people became farmers , sett led i n l arge dense populat ions , and began to f i nd things t o be more i mportant t o them than other people (Montagu : pers . comm . ). Montagu , Boulding , and the others think , furthermore , t hat Lorenz 1 and Ardrey ' s views are dangerous . They justify the status quo : Wha t we are unwill i ng to acknowl edge as essentially of our own maki ng , the consequence of our own diso r dering i n the man- made environment , we saddle upon Nature , upon "phylogenet icall y programmed" or "innat e " factor s . I t i s very comforting , and i f , somehow, one ca n conne c t it a ll with findings on greyl ag goslings , s t udie d for t heir "releaser me chani sms ," and relate the findi ngs on fish , bi rds , and otrLer animals to man , i t makes everything a ll the easier to understand and to a ccept (Montagu 1968 :1 6 ) ; Nothing coul d more effectively prolong man ' s fight i ng behavior than a belief that aggression i s in our genes . An unwelcome cultural inheritance can be eradicated fairly qui ckly and easil y , but the incent ive to do it is l acking while peopl e believe that aggression i s innate and i nstinctive with us ... (Carri ghar 1968 : 50 ); I dealism of some sort i s ne ce ssary for j u t i ficat i on and l egi t imation , for no l ine of policy c an be pursued for very long without self- justification . A line of argument like that of Ardrey ' s , therefore, s eems to leg i t i mate our present morality , i n regarding the threat system as dominant at all costs , by reference to our 45 b i ologica l ancestors . I f t he names of both antiquity and of scie nce can be drawn upon to legitimat e our behavior , the moral uneasiness about . . . Viet nam may be assuaged (Boulding 1968 : 89 ). Ardrey and -Lorenz want to strip culture from man and say , This i s what natural man i s like . Mont agu and the others po i nt out tha t this i s i mpossible : culture is the essential characterist ic of the nature of man . But we cannot know what natural ma n would be like , first because culture has been a round for so long and is t he crucial factor in human biological evolution ; and second , because human infants cannot survive without other human beings to take c are of them- -in ways they learned from their parents . The influence of culture begi ns at the moment of birth , and it can be a r gued tha t it begins even before , s i nce the unborn child perceives light and dark , sound , and motion. War i s no t natural , but a cultural invention that we expect to reso l ve our c onflicts . It is an idea that spreads from society to society and has an apparently irresistible qua lity : those people that do no t accept t he i dea must flee or probably be eliminated culturally and even phy s ically although the effec- tiveness of non-violence has r arely been te s ted . The usual a lternat i ve i s to borrow the i dea, i . e ., fight back. The dispute over nurture versus nature c ontinues to rage and probably will never be put to re st . On the con- t rary , more books and a rticle s appear every month using the ethologica l or bi ologica l paradi gm . The following 46 quotations represent what the nature movement started out to be, and the illogical extreme that it has become . When ethology f irst began to make an i rrpact i n England and America it was like a breath of fresh a ir blowing through the laboratories where rats turne d t re admills , threaded mazes , and pressed levers , and psycholog i sts constructed theori es of human behaviour therefrom . We (we non-bi ologi sts , that is) l earnt from the ethologts ts that i t was unscient i fic to generalize even from one breed of rat to another , let a l one from r ats t o human beings ; and we began to get an i nkli ng of what the per- ceptual and sensory worl d of other species than our own might be l ike , and how anthropomorphic it could be to t a l k of animals " thinking" this or "deciding" to do that (Anon . 1 972) . From such measured science, the mad rush to the side of nature i s epitomized in the following , a l e t ter by a geog- rapher to the editor of Natural His tory on Marvi n Harris ' (1 972 ) explanation for warfare , discussed l ater . While it may confuse the reader a bit , I ir.clude the letter here as the most appropriate spot . Although I f i nd Marvin Harris t s ideas on the orig i n and function of warfare ( "Warfare Old and New ," March , 1972 ) fascinating , I also f i nd them highly speculative . Harris ' s conclusion--that warfare i s not associated with ins tinct --i s unfounded . Certainly , he fails to "prove " his conclusion . Since Harr is objects to the view that we go to war because of our aggress ive ani mal ins tinct s , I recommend he delete the ·word animal s o that it reads "because of our aggressive i nst incts . 11 He may also wi sh to delete the term ~gressive because humans also be come i nvolve d i n warfare for protective reasons e However , the "urge " to protect ourse lve s , our posses- sions, intere s t s , territori es , and beliefs is something we are born with ; it i s purely inst inctive in origin . Because it is , we pl ace so much value on protect ion (or se curity ) that we be come aggressive about it and employ jus t about every cultural means--i ncluding bombing raids halfway around the g lobe- -in an eff or t to guarantee it . Not only warfare , but probably everything we do- - biting our nails, mowing the l awn , or holdi ng up a bank-- i s at 47 least to some degree determined by , or based on, our i nstincts . Careful observation of childreri ' s behavior , for example , suggests that each and every act i v ity (normal) peopl e engage in serves the purpose of sat isfy- ing one or more of. our bas i c needs : the needs for i dentity , se curity , and stimulation (to use Robert Ardrey ' s termi nology ) . Because our instincts are an integral part of us ... , we cannot detach ourselves, our t hinking , or our behavior from them . No matter how often we are told to "love thy neighbor, 11 the overwhelming ma jor ity of humans will remain ego-centric , selfish , and narrow- minded , and groups will remai n ethno-centric . Sermons and philos o- phies cannot change human nature . For this reason , I contend that if it seems that s ome of our a ctions a re not ass ociated with our i nstincts , the likely reason i s that we fail to see t he obs cure conne ctions that (must ) exist be t ween our hid den , i nnate tendencies and our behavior (Reitsma 1972 : 4- 6). While t he letter speaks for itself , permit me to comment on t he obvious : Professor Reitsma demands "proof " from Harris , yet any "proof" for hi s own position is strictly a priori , that of the Compleat Determinist , and not scientific at all . 2 . It is i n our minds . Frustrat i on leads to a ggress i on .--This theory state s that when one ' s hopes , expectations , and needs are not me t or are thwarted , one be comes angry and tries to remove obstacles t hrough the use of violence ; t herefore , one i s also trying to adapt or to cope with the problem at hand . Fa ctors that may i nfluence t he aggressive response are : the nature of the frustration itse l f , previous experi ence , fuaturity , and feel- i ngs of the one being frustrated , and available alternat i ve s t o aggressive behavior (Gilula and Daniels 1969 : 398 ) . I n 48 order Lo re duce violence , frustrations must be removed , e . g ., violations of civil rights , economic deprivat ion , socia l stress , an _ constructiye expressions of aggr es s i on l earne d [ hi tt i ng a Bozo doll, compet i ng in sport s , buildi ng buildings and dams] (Gilula and Dani e l s 1969 : 398 ). There i s another version of this theory which ho l ds that the s ocia lization proce ss by its very nature is frus - trat i ng , i nevitably producing i nternal conflic t and aggres- sion (Holloway 1968 ). Since aggre ss i on i s inevi tabl e , it must be di rected i nt o useful enterpri ses : sports , public works , explorat ion , and hostility channeled i nto literature and the arts . William J ame s ca lled such activit i es "moral e quivalents to war . " Both rendit ions of the frustration- aggres s ion theory s u£fer from an a i lment common to most psychological theories : u.nfa lsifiabili ty . Every instance of violence is expl ained as the result of some frust r at ion , delineate d through the open- ended attribute s given above . This is not to say that we will not become a ngry , enraged , or hi t or kill when we cannot have something we want ; or that people will not try to free themse l ve s from depriva tion , exploitation , and de s pair through violence ; or tha t frustration does not illuinina te a particul a r ca se . I t is to s ay , howeve r , that a theor y which explai ns everythi ng--we cannot a sk what it would be like not to be frus t rated--explains nothing , contrary to the belief of many behavioral scienti s ts tha t the more inclusive their 49 theory i s , the better i t i s (Popper 1965 : 37 ff . ) . Like those who use Freud 1 s theory of sexuality as a unitary the ory , or Ju.ng r s heory of the uncons cious , those who appl y frustration- therefore- aggression across the board are argui ng from an unassailable po s ition . The psychiatrist stresses the similarities , draws the anal ogie s , pre sents the pictures in a new light with new association , as a rhetorical device to focus attent ion on t he patient ' s problems . But his tendency is then to treat his rhetoric as a theory of human development . The plaus ibility and success of hi s rhetoric l eads him to think tha t his hypothe ses of human development have been proved . The the ory is surrounded by a cloud of logica l operations and protecte d from the ne ed for verification which, as a learning theory , i t would re.qui re ( Louch 1966 : 221) . Karl Popper , in 1919 , i nvented an extremely useful principl e , guideline, rule-of-thumb to judge the signi ficance of theorie s labeled as "scientific ." At the time , he was struck by what seemed to be an essential difference between Eins t e in rs t he ory of gravity on the one hand , and Marx 1 s theory of history, Freud 1 s theory of sexual ity, and Adler rs theory of the wil l to power on the other hand , a ll considered to be scientific. He determined the essentia l difference to be that Einste in rs theory could be tested, as indeed i t was by the astronomical observat ions made by Eddi ngton during a solar eclipse--which proved t hat heavy bodies (here the sun) do indeed at tra c t light (of the stars) as well as material bodies (the planets of our sol ar system) . Had Eddi ngton found no differ ences in di stances between stars during s olar day and sol a r night , the new theory of gravi ty would have had 50 to be reworked or scrapped . But Freud, Marx , and Adle r could find no case s i n which their theories did not apply . Fur- thermore , Freud and Adler could and did give radically dif- ferent yet i nternally consistent analyses of the same event . Everywhere they looked , they found confirmation for thei r theories. Therefore, Popper co nc l udes , these theor i es are non-scientific . He di d not mean , however , that be cause t hese t heories could not be tested they had no meani ng . He was not concerned with meani ng at this time . Popper t s verifiability pri ncipl e was adopted and a l tered by those in a philosophical special ty , logical pos- i tivism , that began in the earl y 1920 ' s i n Vienna with phi l osophers Moritz Schlick, Rudo l f Carnap , Herbert Fiegl , and sc i entists and mathematicians . The concern of the Vienna Circle , and what they held to be the proper concern of philos - ophy , was to distingui sh between statements tha t are nonsense and those that are sense , whi le s cience attends to the busi- ness of decidi ng if statements that make sense are true or f alse . To de termine sense- nonsense , they borrowed Popper ' s verifiabi l i ty principle , which he called a "cr iterion of demarca tion, " but instead of using it to determine the s cien- tific or non- scientifi c status of a statement , t he y inter- preted i t so that scientific stateutents were meaningful and therefore worthy of discussion , while non- scient i f ic state- ments were metaphysica l and therefore meani ngless pseudoprob- lems i ncapabl e of solution because they offer nothing to be 51 solved . Popper did not intend, however , that all statements and theories that could not be tested should be junked , or that the line of demarcation in many cases was not blurry . He was trying to provi de some measure of scientifi c status (Popper 1965 : 39- 40) . Above I sai d that the frustration- aggression theory cannot be te sted , i.e . , falsified . Psychologists are qui t e aware of some of the problems with the theory and have tried refinements of the 2 basi c assumptions : that all aggression presupposes frustration, and that all frustration ·1eads t o (instigations of) aggression . Tests of frustration often i nclude adminis t ering el e ctric shocks to people and then observing what they do when it is ' their turn to admi ni ster sho cks ; or tests may be run on pigeons and rats . I n spite of the art i ficial testing s i tuations (outside the psych l ab one woul d probabl y be arrested for sho cki ng peopl e elect r ica lly ) , and psychologi sts 1 admiss i ons " that the level of aggressiqn shown by the frustrated subjects was not very great relat i ve t o the level possible in the s i tuation" (Berkowitz 1969 : 9) , it appears to me that attempts to refine and test continue to be made agains t a background of unfalsifiable assumpt ions . For example , what ki nd of risk is Berkowi tz taking in testing when he calls inhibitions masks of otherwise aggressive reaction to thwarting , and a non- aggressive response t o frus - t rati on a habit? Still , even if they only c onfirm i ntui t ions or common sense, and granting the large assumpt ion that 52 producing frus tration and anger i n the l aboratory through the applicat i on of ele ctric shock i s analogous to frustrations in everyday life in our soc i ety, or that the shocks do i ndeed produce frustrat i on , the psychologists ' experiments are i nteresting . Berkowi tz (1 966 ) has carried out such experi- ments , i n one case on t he effect t he presence or absence of a "st i mulus to aggression ," e . g ., a gun , has on the expressi on of violent aggression . As we suspe cted , t he pre sence of the guns affect ed both the number of shocks t he s tudents gave their partners and how long they he l d t he ke y down for ea ch shock ... from a stati st ica l point of vi ew , our most signi fi cant finding was that the angry men who saw the guns gave more shocks than any other group (1968 Berkowit z has te s te d a corollary hypothesis that violence will not erupt unt il a third factor , in addition t o strong frustrat ion and presence of cues , is present : low i nhib- itions . To what extent i s violence tolerated or en couraged withi n a s ocie ty? He has f ound that rather than having a cathart ic effe ct by subli mating violence and providing an outlet fo r "the spontaneous accumulation of some excitat ion or substance i n neural cent ers " (Lorenz 1966) , the obser- vation of and part icipation in violence encourages and l egitimizes greater violence . In hi s tests , afte r a group of students saw a movie in whi ch they thought the violence jus- tified , they administered more shocks to partners who had previously shocked them (and thereby frus t rated them ) than any other group . 53 The socia l i mplications of the research I have described are clear , though they are much easier to re ci te than to act on . A society that wants fewer violent outbreaks should reduce frustration , leave i nhibitions i ntact and remove immediate cues that can set off aggressive acts (Berkowitz 1968 People learn how to be aggressive .--Social l earning theory states that aggressive behavior is learned , by wat ch- i ng and imitat ion, and does not require frustration for its expression . Reward or punishment of such behavior reflects t he basic values and "adaptive behaviors " of the group . In American culture, whe r e achievement , self- reliance , and i ndividual self-interest are valued highly , we also find a relatively high emphasis on military glory, a re l- atively high incidence of personal crime , and a soci ety characterized by a relatively high degree of belli- cosity ...• From this theory we infer that as long as a nation values and accepts violence as an effective coping strategy , violent ber..avior will continue (Gilula and Daniels 1969 : 398) . I n contrast , the Semai of Malaya conce i ve of themselves as simply not the sort of people who would hurt ea ch other ; they see themselves as nonviolent . Thi s is not merely an ideal to strive for . They do not say "Anger i s bad" or 11 It is forbid- den to hit pe ople ." They say , "We do no t get angry," and "We do not hit people ." The Semai do get angry and quarrel, and they do hit, but not often . They show their ange r not in . violence but in mutual avo i dance and by spreading rumors behind each other rs backs . Any more direct expression of aggression than name ca lling or t hrowing a few object s around i s very rare . They believe it i s wrong to frustrate some- one ' s desires or force a chi l d to do s omething it does not 54 want to do . They have a horror of physical vio lence; one adult Hov.ld never hit another be cause "suppose he hi t you back?" Nor shoul d one - hit a child becaus e, "How would you feel i f it died? '' (Dentan 1968 : 55- 58 ) . We can i dentify with Berkowitz i n t his case and say that the Semai experience little frustration , that they encourage inhibitions against violence , and that they do not live with aggressive cue s . While .Americans are ambivalent about violence and devote cons iderable effort to explaining and justifying it , the Yanomamo of southern Venezuela do not . I nstead , they have elaborated waiteri (ferocity) into an ideological complex- -a male complex . The socialization process sele cts for and encourages ferocity . Masculinity and aggressiveness are instilled i n small children from an early age . It i s common to see parents teas e a small boy to strike at his tormentors , rewarding his anger with approving l aughter . Girls , on the other hand , are taught to acquiesce timidly to the punishment they receive from their brothers, so that by the time children are six or seven years old, the boys have a lready learned that it is appropriate to bully the girls and spend a great deal of time at mischievous pranks calculated to intimidate them . Boys ... are encouraged to be fierce fighters . They have numerous opportunitie s to participate in fights .... They are pres sed into the fighting by their adult superiors , but are given privileged positions in raiding parties until they acquire the necessary skills and experience .... Us ually a boy does not take an active ro l e in raiding until he is seventeen years old , and even then he may be s o frignte ned t hat he will fake illness and return home before the enemy village has been reache d . Yanomamo boys , like a ll boys , fear pain and persona l danger . They must be forced to tole rate and l earn to a ccept ferocity as a way of life (Chagnon 1968a : 130) . 55 That way of l ife i nclude s gi vi ng and a ccept i ng cha llenges , temper tantrums , wife-beating , beating contests wi th other males , a nd war expeditions . Yanomamo enculturation of bellicosity i s one of the be s t documented events i n rece nt ethnography . It helps to support the fo llowing gener a lization by Andre ski . I n every warlike polity ... there are elaborate social arrangements 1vhich st i mulate martial ardour by playing upon vanity, fe a r of contempt , sexual desire, filia l and fraterna l attachment , loyalty to the group and other sentiments . It seems reasonable to suppose that if there was an innat e propensity t o wa r-making, such a stimu- lation would be unnece ssary . If human beings were in fact endowe d with an innate proclivity for war , i t would not be ne ce ssary to i ndoctrinate them with warlike virtues; and the mere fact tha t in so many societies past and present so reuch time has been devoted to such an indoctrination proves that there is no i nstinct for war (196 8 : 187) . Thi s forms part of the basis for certa i n questions for which data have been collect ed i n t his study . Again, I am not willing to make a direct causal connect ion between t he pres- ence of certa in pra ctices and value s re l a ted to warfare and the frequency of warfare experienced by a particular society . But the presence or a bsence of these values I cons i der to be vitally i mportant with regard to tho 0 e reasons , i ntent ions , a nd purposes of an individual anonymous warrior . One cer- t a i nly does not want to c ommi t the f a lla cy that 11 the Yanomamo go to war because they are warlike , 11 nor to be merely offer- i ng a triv i a l truth that people will s eek to a chieve what they value . I ndeed , the l a tt er may no t be so tri vial after all fo r t hose of us who have grown up i n a culture that the 56 psychologists say values v iolen ce but whose politician s in e very ot~er breath use the words peace , re co nciliat ion , generosity, mercy . The prob lem may lie in what const itutes v iolen ce tJ the members of a particular ~ocie ty . Unfortu- nately , the questionnaire question for this study which attempted to get at the way violence is va l ued in a particu- l ar society neglected (or I was not able to i nvent) to con - trol for an etic-emic distinction . Answers to that particu- lar question were coded impression istically in most cases ; occasionally an ethnographer would help by giving clear evi- de nc e for a particular answer . While spe c ific details of questionnaire construction and rationa le properly belong in Chapte r II , in a crude way this study wil l end up testing many, if not all , of the explanations offered for the oc cu r- ren ce of warfare discussed in this es say . In any case , I do have on e piece of evidence that there exists , at least in our own culture , dramatic variation- -according to subcultur e --of what defines or identifies an act as v iolent . The Institute for Social Research of the University cf Michigan conducted i ntervi ews i n the summer of 1969 with a nation- wide representative sample of 1 , 374 American men , col- lecting attitudinal data on what acts they considered as vio- le n , what they thought causes those a ct s , and h ow they thought violence should be dealt with . The results were iJluminating . I do have a major o bjection to the study : why , especially since the project was headed by a woman 57 psychiatrist , didn ' t the r esearch workers collect comparable data for 1374 American women at the same time? I n terms of the sociology of knowledge , it looks as though the research workers were guided by an implicit value : women are not re l- evant to v i olence , or that a violent act is a male act . Anthropology of psychology as i de , " one of t he key findings of the study was that what people mean by violence has a great deal of significance for other attitudes . And t he more one l abe l ed an a ct as vio l ence , the more fo r ce one was will ing to unl eash to combat it (ISR 197 1 : 4) . As support for my own philosophy of language pos i tion on definitions , the study found that generally the men all knew what violence was : " ... something bad , worthless , fier ce , strong , and unne cessary . Blacks were the onl y except i on to thi s ; they tended to define violence as weaker , le ss bad , and l ess fier ce than other groups " (1971 :4) . Yet 645 men of the sample did not agree on which acts are violent a cts . In a gross breakdown of events c~nsidered to be violent (Table 1), looting was ment ioned by 85 per cent of the i nformants , burglary by 65 per cent , draft - card burning by 58 per cent , police beating students by 56 per cent , police shooting looters by 32 per cent (ISR 1971 :4) . In answer to the ques- tion of how burglary , a crime agai nst property , and draft- card burning could be considered s i milarly violent , The study directors believe that one answer lies wi th the illegitimacy of the acts . Men tended to put under the rubric of violence actions they categorized as illegitimate , a.cld the majority of American men believed 58 1'ABLE 1 • . WHO CALlS WHAT VIOLENCE* College White Students Union Bl acks Members POLICE Police beating students is viole nce 79% 45% 82% Police shooting looters i s v i olence 43 23 59 Police fri ski ng is violence 1 6, 10 34 BURGLARY Loot i ng i s violence 76 91 74 Burglary i s violence 47 67 70 DISSENT Student protest i s violence 18 43 23 Sit-ins are vi olence 4 24 1 5 Draft-card burni ng i s violence 26 63 51 Denial of civil right s i s vio l ence 54 40 70 N ( 63 ) (279) (303) *I SR 1971 : 4 . 59 that the state has a monopoly on legitimate force ( I SR 1971 : 5) . '11he study found that tpe men were more willing to use for ce against individuals or groups they did not like , and that they did not c onsider for ce to be violence . So , police a c tion against a disliked group , no matter how much physical damage they inf licted , woul d not be considered as vio l ence . The research workers tie t hese findings into i n group--out group behavior , dis cussed elsewhere in t his e ssay under the emi c statement , "It is wrong to kill a human being but a ll right to kill an enemy ." The research workers also i nter- viewed the informants about 5 values related to the legiti- macy of force : retribut ive justice , self- defense , person versus property , liberalism versus conservatism , and kind- ness . They found that retributive justice--"an eye for an eye " --and self- defense were t he most potent values among American men generally in determini ng attitudes toward violence--ki ndness was the least potent : 11 0ne i s tempte d to say that the values that justify v iolence are more ' i mportant in the determination of attitudes than the values that oppose it . 11 The study f ound that the more j ustifiable a man found the use of vio l ence by police to control hoodlums , students , and rioting blacks , the more he believed in the value of self- defense , retributive justice , material over humanis- tic values , property over persons , and the l ess he beli eved i n kindness (I SR 1971 : 5- 6). However , t he more a Black American held the values of self- defense and retri butive justice to be i mportant , the more he believed that the goals of student demonstrations and pro- test s would make a better world , that police were not 60 trustworthy , and that police a ct s of shooting looters , beat- i ng students , and fri ski ng were ones of violence . A more hopeful :finding was that t he more education a man had , the les s he approve d of police violence for social control , the more he was able to i dent i fy with members of an out- group , and--of special i nterest to me--the less he valued retri butive just i ce (ISR 1971 : 6) . Obvi ously , the attitudes toward violence documented for our own society break down s ignificant ly into sub- cultures : those of college students of 1969 , white uni on members of 1969 , and Blacks in 1969 . With even f iner dis- t i nctions , other i nteresting percent ages mi ght have turned · up , but the sampl e s i ze would dwindle . I n any case , what t hese psychol ogists were l ooking at were values shared by and probably learned wi thi n a particul ar group of people ~ Maybury-Lewi s (1 967 : 305- 307 ) suggests a promising hypothes i s that I was not able to collect data for i n t his proje ct but i s probably, I think , very signi ficant in mat ters of at ti tude , bellicos ity , and war frequency . I n compari ng Ge- speaking groups in South Ameri ca--the Shavante , Kayap6 , Sherente , a nd Tirnbira-- 3 were troubled with severe factional- i sm , frequent killings withi n and between communitie s , and break- up of communities . Factionali sm is l ess severe or absent among the Eastern Tirnbira . All 4 societies have age- set systems , age- moieties for log race s , and neither the Timbira nor the Kayapo have clans or lineages (potent i a l 61 units of competit i on) . But while the Timbira are internall y harmonious , the others have continual i nternecine warfare . The one institution found in the fractious 3 but not among the Timbira is the men rs house . Rather than being a means to I achieve harmony or Tiger ' s stable backbone of society, the existence of a men ' s house has just the oppos i te effect . I n the men ' s house s of the Central Ge, boys are taught not onl y ceremoni al and fraternal dut ie s , but the values of manliness . One of the ways manliness is expressed i s i n bellicosity , expressed ritually by initiated male s against uninitiated ones , and i n the most i mportant initiation ritual of all , ceremonial gang rape of select (and terrified ) women, the ultimate outsiders . Bellicosity is not conf ined to ritual situations but spills over i nt o secular life , Maybury- Lewis argues , stimulating political fa ctionalism , often violent , within communities and , as communi ties fission , between communities . The Timbira not only do not have a men ' s house , they also do not have a violent or possessive attitude toward women : sex is easy and fun . There are several all-male cor- porate groups but a l so many community a ctivities with opposite- sex part icipation . The arts of diplomacy, compro- mi se , peace- making , and generosity are those valued by the Timbira . Maybury-Lewis' speculation about the i nf l uence of the men ' s house shou.ld be readily te s table . The Yanomamo do not have a men ' s house , but the village is so constructed that i t could be described as a s i ngle huge round house with 62 c ompart□e nts for i ndividual families , each compartmen t open to the ce nter plaza and l argely open to it s neighbors . The point i s , men can read_i l y gather in any spot for their drug- t aki ng or pal avers , and their behavior i s constant l y i n the public eye. Male s learn from and t each other mal es . Their education and a chi evement are continually assesse d by other men . 3 . Cul tural determinism and war . I do not know what is true . I do not know the meani ng of the uni verse . But i n the mids t of doubt , in t he collapse of creeds , there i s one thing I do not doubt , that no man who lives i n the same worl d wi t h most of us can doubt, and that is that the faith i s true and adorabl e which l eads a s oldier to throw away his life in obedien ce to a blindly a ccepted du.ty , i n a cause which he little under- stands , i n a pl an of campaign of which he has no notion , under tact ics of whi.ch he does not see the use . --Oliver Wendell Holme s , Memori a l Day Address , Harvard , 1895 One of the more recent and faster spreading s chools of explanat ion i n regard to warfare i s that of econowics and ecology , a nd for t he following discuss ion I am concerned with the positions and contributions of Whi te , Sahlins, Harri s , and Divale . Previous discussio n has deal t with the i ndi v id- ua l --in biol ogy , evolutionary hi s tory , and psychology . But reducing explanat i ons of a group activity (warfare ) to the l evel of the i ndi v i dual i s h i ghl y di sapproved , r i ski ng charges of "reductioni sm" or , even worse , "psycho l ogi sm . 11 What Harris is calling for i s a revival, more s ophisticated , 63 of course , of the 19th century search for l aws in the soci o- cultural ~ist ory of manki nd . I t is a back swi ng of a pendu- l um , the anti thes i s of; the historical ( particul arist) , structural- funct i onal , and di ffusionist s chools of anthropol- ogy that deve l oped during the f i rst hal f of thi s century . They we r e not onl y not concerned wi th t he di scovery of so cial l aws but thought s uch di s covery and l aws i mpossible . The i r concern was wi th the i ndi vidual pers on , t he indi vi dual cul- ture , and the vari et i es r ather t han the uni f ormi t y of pat - tern . They were i nterest ed i n t he hi st or i ca l deve lopment of i ndividu~l so cieties and historica l relat ions betwe en s oci- eties , and i n the non-historica l s tructure and worki ng of soci ety , rather than determini sm and Laws of His tory . Mos t of the really good de s cr i ptions of warfare among t he I ndi ans of North America were done by the American hist orica lists , a nd t hose f or Africa by the Bri tish structural-f un ctionalists . Anthropol ogi s t s have never a ctually l os t t he i r i n t erest in r egul ari ties , cont rary to Harris ' charge , a nd t he i nte r e st i s even stronger , threatened as we are with i nundat i on by our accumulated data . But Harris is crit ical that sci ent i sm i s not more widely f ollowed , spe cifically that our search f or regul arities i s no t concerned wi th c ausality and or i gi ns . Various strategems have been introduced that avoid state- ments of causality , while conveying the impression that an explanation i s be i ng offered .... We have so-called functional explanations ; we have correlat ions i n wnich it is not known how the c ausal arr ow points , and we have "a ccounting" for in terms of paradigmat i c cognitive frames , whi ch are a ccepted as givens , a l though nothing i s 64 kriown about how long they have existed (Harri s 1968 : 2) . Harri s ' pr imary reason for writing ~~he Rise of Anthropol ogi- cal Theory i s "to assert the methodological priority of the search for the laws of history in the sci ence of man" (1 968 : 3) . He regards mi ddle- range theori es a s e cle cti c, chaotic , usele ss , and based upon t he unrealisti c expectat i on that eventually i t will a ll make sense . He contends that anthro- pology needs a general t heory of hi story i n order t o make decisions about how research funds are to be spent , and decis ions i n social engineering , especi a lly of i nternational deve l opment programs . The ult i ma ratio of socia l u seful ness for a scientific paradigm i s a crucia l i ssue i n both profes- s ional ethics and the philosophy of social s cience . I ts philosophical i mplica tions will be taken up later . Et i cs , emi c s, and l aws .--The basic analytic s caffold for the fo llowi ng di s cussion i s the etic-emic disjunct i on . Or, the one hand , it i s useful as heurist ic to cope with the t wists and tu.rns of sociocultural theory . On the other hand , i t is a speci f ic i ssue in the philosophy of anthropology , it s formul ation i ndigenous to the di scipline . Within cultural anthropology , l inguistics has often been l auded as the mos t rigorous, theoretically fruitful , and therefore suc cessful of the spe cial ties . As a consequence , investigators of socio- cultura l phenomena have tried to emulate this success by borrowing the basic analytic fraii1ework of descriptive 65 linguistics and applying it to human behavior . Many problems have ari s en in this transfer , and Pelt o ( 1970) de clares that the debate between emi'cists and etici sts i s fundamental . Briefly, what is going on? Etics in linguist ic s is specific and limited . It refers to a universa l grid describ ing the production of spee ch sounds so t hat the sounds of any language re c orded by any linguist can be reproduced on sight by any other lin- guist . I t isolates discrete units of sound i n the spee ch st ream and describes the units with varying fineness of detail i n standard symbol s . Theoretically, the grid shoul d i nclude all possi bl e sounds that the human spee ch me chani sm can make . Mechanism i s an appropriate term ; phonetic description is mechanical . Nevertheless , it i s subject to t he vagaries of the linguistts ear and the informant t s compe- tence . Di sagreements between investigators occur a t even the lowest l evel of analys is . Once the gri d has been used t o i solate and describe sounds , analysis shift s diametrica lly from that of a universal open system to that of a specifi c closed system-- emi.cs . One concludes with a statement a bout t he significant structural features within a l anguage . Yet one cannot understand one side of the disjunction without the other : one cannot make mutually exclusive categories of sounds without the sounds themselves ; sounds left i n iso- l at ion are meaningless . The principle of opposit ion is operating : dark cannot be perceived unless you know what 66 light i s ; truth cannot be recognized unles s you know what falsehoo d i s . 3 In the transfer to sociocultural phenomena , the etic- emic dist inction is similar to Robert Merton ' s concept of real and ideal cultura l patterns . The etic is how people a ctuall y behave in the j udgment of the anthropologist ; the emic is how the people themselve s conceive of their behavior--their reasons , intentions , and expl anations . Etic and emic descriptions are often contradj_ctory. The anthro- pologist Is and informant 1 s explanat ions for a s i ngle event may be complet ely different and each deny the validity of the other ' s explanation . But Harr is insists that the et ic is prior : There is no error more common or devastating than to con- fuse what people say , wish , dream, and be l ieve they do wi t h what they actually do ( 1971 : 149 ; ital ics a dded ) . The anthropologist is to lay an "etic grid " or macro- theory-- i n Harri s ' case that of cultural materialism--upon the cul- ture under investigation and isolate the dis crete units of behavior, anal ogous to phones , which subsequent l y are organize d i nt o emic categori es . But somewhere along the way what is heuristic in lingui stics becomes the real world in culture , illustrated by Harris ' emphasis on what people actually do quoted above . How transferable is the linguistic paradi gm? An etic grid says nothing about the relationshi ps among s ounds--that comes with internal ernic analysis . But Harri s ' analogous "etic " uni ts--basic resource s , tools , 67 techniques of production , energy , supplie s , and control--are i nextricabl y interrelated . Furthermore , etic gri ds are no t causal, but cultura l materialism is . White , Sahlins , Kapl an , and Harris represent the etic po i nt of view and place them- selves i n opposition to tho se anthropologists , spe cifically Boas and his students , who stand on the emic side . Thei r etic gri d i s that of cultural materialism and evol ution . Before dealing directly with these theorists , however , I want to digress a bit and talk about Levi-Strauss ' use of the etic-emic disjunction in analyzing soci ocultural phenom- e .. a , be cause his more literal transfer of the paradigm pro- vides comparison and contrast for what the others have done . Levi-Strauss ' paper, ''Structural Analysis in Li nguisti cs and Anthropology " (1945) , is an extraordinary attempt to apply linguistic method to the analysis of social relationships , to kinship systems in particular , i n order to provide expla- nations as to how and implicitly why these systems can be arbitrarily derive d and yet function with regularity and "effect iveness " i n maintaining society . As lingui sts are no longer i nterested in treating terms within a l anguage as independent but as elements in relationships that form a sys- tem to be subsumed under general laws , so the anthropologist, Levi- Strauss argues , should view what ·we call culture t raits as elements within a social system . Kinship terms are analo- gous to phonemes , and their arrangement analogous to phonemic systems . Levi- Strauss means somethi ng more than analogy , 68 however . The structure of language i s evi dence of inherent structures of the human mind . Therefore , since ki nship sys- tems are another manife s tation of that mind , they must have structure a l so . I ndeed , they must have the same structure as language . At the time he was writing , attempts to apply struc- tural lingui st ic method to social data we re apparently unsat- i sfactory ; the results of such analysis were more complex and less elegant and explanatory than the raw data on which they were based . Why? Because kin terms were treated as words , between whi ch there is no necessary relationship , rather thari broken down into nphonemes . 11 To obtain a structural law , the linguist analyzes phonemes into differential elements or dis - tinctive features (etics ) which he then organizes into "pairs of oppos itions " (ernics) . The s e elements and their relations exis t independent of psychological , natural, and physical factors . To not only demonstrate but to prove hi s thesis , Levi- Strauss applie s formal structural lingui s tic analysis to the social relat i onship between mother ' s brother and sister ' s son and presents us with a general law : The relation between maternal uncle and ne phew i s to t he relation between brother and s i ster as the relation between father and s on i s to that between husband and wife ; if we know 1 pair of relations, ·we can i nfer/predict t he others , since in ea ch of the 2 gener- ations there is always 1 positive relationship and negat ive 69 one . The premi s es from which this law is derived i nclude a brot he r - s i st er- sister ' s son bas ic unit of kinship , the incest taboo, a nd a llia nce theory . But these premises cons i s t of unfalsifiable a ssump- tions . Levi-St rauss , unlike t he descriptive linguist, per- sists in going beyond form--items and their arrangement--into meaning and function . Although recognizing that form does not correlate predictably with meaning , that the symbols or kin terms are arbitrary , he still presents what he c onsiders a predictive model that i ntegra tes form, meaning , · and func - tion . In this paper , and throughout most of his work , he relies upon case i llustration to establish a general law, a practice derived from Durkheim , who believed that 1 good case supported a universal explanation . Quantification i s not considered necessary because the structure of the human mind i s a universal , like Freud ' s structure of the psyche . From what one may call historical i dealism , or aver- sion thereof , l et us move to the etic appl ication of hi stori- cal materialism. The argument of the cul tural evol utioni sts t hat ma teria l conditions are t he caus e for certain things happening in human history is powerful . We are probably more likely to nod our hea ds in agreement while reading analyses of event s in the se terms than when reading Levi-Strauss ' interpretations, although both are seduct i ve albeit mutua lly exclusive a r guments . Let us look more clos ely at thi.s 70 feeli ng of agreement , and. then look at a ctual cases in warfare studies . I n their concerh with methodological rigor, anthropol- ogists often insist on drawing t i ght distinct ions between hypotheses, theories , and laws, especially those anthropolo- gi sts such as Manners and Kaplan (1968) , influenced by the l ogical positivi sts . Nielsen ( 1967), a philosopher of another school, offers an a lternative that is no le ss demand- ing but i s more in accord , I think , with the ways in which human be i ngs ask questions and seek answers . What hypotheses and t heories do i s to make sense out of a phenomenon : a hypothesis "closes up a gap of puzzlement "; a theory is a wider sort of explanation which "enables men to set their mi nds at ease about a broad range of diverse facts .... " (Nielsen 1967 : 32) . While i t is difficult to make a distinc- tion in all cases , a law does not expl ain anything but expresses the regular mode of action to be expected from a certain class of phenomena or the regular mode of action of a force (Nielsen 1967 : 32) . With the exception of a reference to Lotka in White (1 959), we do not refer to Darwin ' s theory as a law . Instead , Darwin ' s theory has "set our minds at ease " about facts from geology , paleontology, embryology , and biology . That is, fossil and species variability i s made intelligible-- we can look at these phenomena without feeling that they are inexplicable . Thus , 71 ... a theory i s an idea or a manageable handful of i deas by reference to which men can throw light on a l arge class of physical f a cts (Ni e l sen 196 7 : 32) . I n comparing 2 _specific the orie s , Einstein ' s theory of gravitation and Darwin ' s t heory of evolution , there is a striking and significant difference between the two : the fo rmer i s predictive and testable , the latte r i s not . Had Einstein ' s theory not met the test of predictability , as it did with Eddington ' s findings , it would have faced refutat ion and radical revi sion (Popper 1965 : 36 ). But, as Nielsen points out , theory in biology and the social sciences is more often explanatory than predictive . I t i s not because our standards or methods are somehow faulty , but because physical phenomena a nd phenomena involving living forms are different scientific sub jects . Darwin (and a ll of us) did make predic- tions , e. g ., that foxes who were unacquainted wi th and there- fore unafra id of men when first di s covered would soon be come extinct--which i n f a ct happened . But this prediction was not deduced from his theory ; instead, he made i t by appraising the circumstance s or context with the eyes of years of experi- ence . Had the foxe s not become ext i nct , Darwin ' s theo r y would not have be en refuted . Theory-construction proceeds in a variety of valid , use - ful , and i mportant ways , depe nding upon the kinds of facts under study and the kinds of que st ions me n want answered . In the light of t he se dist i nct i ons it would seem that the dream of axiomatizing biology , psychology , and their sister sciences has been entertai ned without full awareness of the variety of problems the sciences c onsider (Nielsen 196 7 : 38 ). 72 The primary g oal i s intelligibili ty , and biological evol u- tionary theory offers a matrix for explanations of particular events . When we nod our heads in agreement , this means that what we are dealing with i s intelligible . But the et ic anthropologists are seeking more than i ntelligibili ty , or they require that i ntelligibility be validated through pre- di c t ion. And if one demands predi ct i ve power , one must tal k in causal terms . Whil e anthropol ogy has been concerned with the problem of cultura l evolution for over a century now , i nd i vidual the- orists vary in t he degree of err.phasis and e xpl i cat ion . Leslie White has grounded hi s theory of cul ture in physics , and his conception of the evolutionary pro cess in general i s the most deterministi c and the leas t concerned with the indi- vidual . I will go i nto it i n some detail , since i t is the most extreme expression of a theoretica l position i nfluenti a l i n warfare s t udi es . White (195 9 ) offers the unitary theory that everything i n the uni verse can be explained in terms of energy , whi ch is the basic and uni versal concept of s cien ce . The physi ca l ground for the cultural theory i s the Second Law of Thermodynamics , whi ch states that within the closed system of the universe , order i s breaki ng down i nto disorder anu chaos , differences are leveling , towards the final state of equilibrium , or maxi mum entropy . But living things , because they are open systems linked to their sur- roundings , for a while at l east can overcome the entropy 73 produced in the process of l iving be cause they draw on free energy out side themselves . This energy or negative entropy i s used not only to maintain life and bal ance positive entropy , but any excess i s ut i lize d to deve l op structures of greater organization , complexity , and efficiency . Thus , s pe- cie s evolve . I ndividual s , however , eventually are overcome by positive entropy as the ir structures s impl y wear out from re s i stance s inside and out s i de themselves, e . g ., agi ng , chronic di sease , mortal wounds . Thus life and death alike re ce ive their most profound and illuminat i ng defini t ions i n terms of thermodynamics . The mai ntenance of life i s a cont i nuous bal ancing of positive entropy wi th negat ive entropy . Dyi ng i s the lo sing battle to overcome po s itive entropy . Death is the state of maximum entropy , of thermodynamica l equilibrium (White 1 959 : 35 ) . The pri mary source of e nergy for all living things i s the sun , and s i nce s olar energ,y i s relatively boundl ess t he expansion of l iv i ng things i s l i mited only by the capacity of the eart h to "a cc ommodate " them_ This expansion is both quant itat i ve (through reproduction ) and qualita tive (thro~gh the development of higher f orms of life) . Whit e spe cif ie s "higher" as meani ng greater struct ural organizat ion a nd more concentrated energy , e . g . , animals are more highly deve loped thermodynami c systems than plants ; mammals , than r ept i les . I n the struggl e to exist and survi ve , an organi sm a d j usts t o it s habita t in terms of such fact ors as temperature , humidity , r adi at ion , and food , and competes with other organ- isms for habi tat and f ood . 'rhose organisms with the most. 74 efficient 1'energy-capturing device s 11 have the advantage i n this co~pet ition and will hold it . The tendency of the life process i tse lf i s to i n crease the mass of the organic sys- tems , and t he r a te of circulat i on of matte r through the s ys- tem and mat t er ' s transformation i nto energy (Whi te 1959 : 37 ), regardl ess of whether the energy is us ed quant itat ively or qualitatively . Yet this process can cont i nue onl y so l ong a s there i s free matter and energy available . 4 Man i s an organism , therefore he must ad just t o his habitat , compete for that habitat (White uses the phrase "defense from enemies n), and reproduce . To do this , man must capture and utili ze energy, whi ch he a ccompli shes t hrough hi s body and through culture . Culture i s both a means by whi ch ,;-re capture and u se ene r gy , and the product of that use . "Cul ture " is but t he name of the fo rm i n wh i ch the life force s of man as a human bei ng find expres s i on . I t i s an organi zation of energy transformat ions tha t is depend- ent upon symboling (White 1959 : 38 ). As the f undame ntal process of the human organism is to cap- ture and utili ze free energy , so the function of culture is "the harnessing of ene rgy and putt i ng i t to work i n the serv- ice of man .n Culture cannot exist without man , and it is a l ways found with man . But from t he standpoint of s cientific ex lanation of cul - tura l diversi ties and rocesses of chan e but not of the nature of culture in eneral culture ma be treated as i f it had an existence of i ts own inde endentl of the human species . Ita lics added . The "as if" factor does not render explanations made on the basis of thi s assump- tion f ictitious or nonscient i f ic . The s cience of linguis- t ic s proceeds upon t h i s assumption , and i t i s t he closest 75 approximation to a mature sc i ence that we have on the l eve l of hwnan affairs . Man , the human spe cies or human organism , i s irrelevant to the s cience of linguistics . He , or it , i s likewise irrelevant to the s cience of culture (White 1959 : 15-16) . Since culture may be treated logica lly "as a distinct and autonomous ki nd of system ," and s i nce cultural systems are materi a l systems , ... we may interpret the evolution of culture i n terms of the s ame principles of t hermodynami cs t hat are applicable to biological systems (White 1959 : 39 ). Subsequent ly , White sees cultural systems using the energy t hey ca pture to e xtend themse l ves quantitat i vely ( population growth , group fission) and qualitat ively (higher forms of organization and greater c oncentrat ions of energy ). As t he amount of ener gy harnessed by s ociocultural sys- tems incr eases per capita per year , the systems not on l y i ncrease i n size , but be come more highly evolved , i . e ., they become more differentiated structurally and more specialized functionally (White 1959 : 40 ). I n "dis covering" cultural structure and function , one does not need to consider the environment at all ; such cons i derat ion be longs t o the particular case : the law of f alling bodi es i s valuable precisely because it ignore s the i nfluence s of atmosphere and the composit ion and structure of the falling body . I n exa ctly the same way , the culturologist i s t r yi ng to formulate laws of behavior of cultura l systems . Like the physicis t, he wants vali d uni versal s . I f one wishes to deal wi th particulars , with particula r cultures or par - ticular fa l ling bodi es , then allowance must of course be made fo r part icular condit i ons in each instance (White 1 959 : 52 ) . How does warfare f i t i nt o this theoreti ca l mat rix? Contending in his earlier work , The Science of Culture 76 (1 949 ), that one cannot understand warfare by looking at the psychological motivat i ons of individuals, Whi te i nsists : Wars are fought between so ciet ie s , between socio cultural systems , betwee n t·ribes and nat ions . It i s the culture of a ny gi ven situat ion that determi nes whe ther warfare shall be engaged i n or not , and i f so how , with whom and for what (1 949 : 131- 32) . Hi s discussion of the problem i s brief , wi th the convict ion that the i ndividual s i nvo l ved are s i mpl y not relevant . In 1959 , while his discussion of war i s again brief, White give s a more sympathe tic picture of man as vict im , especi all y of the agricultura l revolut i on , the deve lopment of so cia l s t rat - ificat ion , a nd the concept of personal property . An elaboration and di re ction applicat ion of White ' s theoretica l pos i t ion i s contained i n two papers by Newcomb dealing specif ically wi th warfare . The first (1 950) i s a re- exami nat ion of the causes of warfare on the Great Pla i ns ana l yzing warfare as i nfluenced by the horse and gun complex , the fur trade, and the dislocat iorn and migrations of nat i ve pe oples as the r esult of white i ntrusion . The anal ys i s appears sound , and was i n fact anti cipa ted by Oscar Lewi s ' study of the effects of the fur trade on Blackfoot cu l ture in 1942 of which apparent l y Newcomb was not aware . The anal ys i s is i n terms of the Chippewa doing somethi ng , the Dakota doing something i n r esponse , and so on . Fo r example : By 1650 the Chippewa and associated tri bes had come i nt o t he territory west of Lake Huron , and by the middl e of the next centur y were we s t of Lake Superior . I t was the mi sfor t une of the Sioux , who had been i n possession of t he Minneso ta territory , to come i nt o violent collision 77 with these invaders equipped wi th the steel kni fe and musket . After a war which l ast ed many decades , the Sioux were for the most part crowded west of the Mississippi . The o ·ibway , however , had had to eje ct the Sioux from Minne s ota because they thems elves were being harasse d by t he oncoming white· settlers ( 1950 : 322) . · It appears that he i s descri bi ng human act i ons , reasons , pur- poses , and s o forth . I n conclusion , he gives the causes of Plains warfare : mi grat ion onto the Plains from east and west , compet ition for the horse, competition over hunting terri.tories and de creasing game , competition for guns , and European politica l machinations of pl aying tribes off against each other (1950 : 327-28) . Newcomb concludes : I n broad te rms we may say that from the i nt roduction of the horse until t he extinction of the bison herds the Plains pe oples were maki ng cont inual and s uccessive adjustments to t he forces of Euro pean culture (1950 : 328 ). While t his theoretical generalizat ion produce s a mild let down a fte r the previ ous ad hoc explanation , t he real diffi- culty comes in t ryi ng to plug this a ccount of Plains history i nto t he proposition t hat ... war will be treated as a type of armed conflict that takes place between societies , meeting in competition for anything that is value d by the groups involved, us u- ally consi s ting of te rritory or certai n products of this terr itory , such as good hunting grounds , oil-producing or agricultural l ands . I t must be emphas i zed tha t this definition says nothing about individuals , for i n terms of this analysis warfare i s held to be a function of s ocio-cultural sys tems , and i ndi viduals are re arded as be i n no more than t he means through which these systems attain the i r ends 1950 : 317 ; ita lic s added) . 78 Or that The motivation of the i ndi vidual i s not the cause of war - fare , i t i s rather the method b which a cul tural irri- tation or need i s satisfied 1950 : 320 ; italics added • Or finally that Plains tri bes did not habitual l y e~gage in war be caus e individual men were "warlike ." I ndivi dua l men were war- like because their socio-cultural systems obliged them to be . The i ndi vidual attitude of war was an expression of the socio-cultural process , by no means its ca use or i nitiat or . Whether i ndivi dual men will fight for obs cure economic reasons , which benefi t t he society and t he i ndi- vidual only i n passi ng , i s unknown . The fa c t of the matter i s that most i ndividuals everywhere , i n all ·war- ring cultures, fight because of i mmedi ate , personal reasons . They fight for glory , for social prestige , to escape civilian frustrat ion , or fo r other i ndividual reasons ; not primarily , or perhaps even knowingly for i mpers onal, broadly cultural reasons . I t s eems probable that men everywhere fight bet t er if they are fighting for personal reasons . Broad cultural ends are more quickly attained by the psychologica l provision of adequate moti- vat ion . Yet i t does not matter f or what reas on the indi- vidual thinks he is fighting and dyi ng , as long as he is sat i sfying the needs and i mperat ives of his culture (1 950 : 329 ; ita lic s added) . While it is laudable that Newcomb should criticize and offer alternatives for the previous superficial and taut ological expl anat ions offered for Plai ns warfare, the ethnographies contai n accounts showing i ndividual s , i n varying degrees to be sure , quite aware of so- ca lled economic or cul tural i mpe r - atives for warfare . The emic i s in accord wi th the etic. In a second article (1 960 ) Newcomb discusses the cul- t urological vi ew of warfare i n general and this t i me offers a typology rather t han a case analys i s . While the paper i s a l so an expansion of background arguments , especially agai nst the utility of psycholo6 ical explanations , in considering the 79 "entire knoKn spectrum of human cultures " Newcomb fi nds four types of co~flict , each associated with and determined by particular te chnologies . Type 1 includes food gatherers living i n ·i solated (e . g ., Australia) or unproductive (e . g ., Great Basin) regions, who tend to be peaceful because there is lit tle of economic value for the small scattered groups to fight over- - no one can accumul ate surpluses , control l arge te rritories , profit from captive l abor , spare much time from t he food quest , and organize i nto large military units . Infringement upon another band ' s territory i s perhaps the most important source of conflict . But conflic ts are generally so infrequent, bri ef , unorgani zed , and i nvolve so few individuals that they must be cons i dered a dis- tinc tive fo rm of warfare (1960 : 327) . Type 2 is what we have i n mind i n referring to pri mi- t ive warfare . People subs i s t primarily as foo d-colle ctors who hunt, or garden part or even most of the t i me . Their mode of production cannot support cities , social stratifi ca- tion , or states . Conflicts tend to be uneconomi c , crude , , sport- like , unorgani zed , and brief (1 960 : 328) . They do not have , i n Malinowski 1 s phrase , "culturally constructive sig- ni ficance ," i . e . , they are not means to enlarge the e conomic base because one ' s neighbors do not have much wealth. "Pr i m- itive warfare i s best comprehended as a transitional type of conflj_ct--transitional between an ordinarily peaceful state of affairs and serious , deadly , compet itive stri fe " (New comb 1960 : 3~8 ) . People carryi ng out primi tive warfare cannot raise or contro l large f i ghting groups , and those groups that 80 are raised cannot stay in the field for very long . Neverthe - l ess , .. • the technologica l l eve l of the se cultures is not so restrictive i n its effect as to preclude war . They are probably more frequent l y in conflict over favored hunt i ng and f i shing grounds , and other natural resources than are the te chnologically mos t pri mitive cultures (1 960 ; 329). Cultures i n Type 3 are based on food produc t i on , which can support l arge dense populations , specia lizat ion , ci ties , and states . Property replace s ki nship as the bas i s of socio- economic organization . Consequent ly, war is highly profit - a ble, and the end of i nc reasing movable and i mmovabl e wealth i s accomplished by means of mobilizing l arge sectors of dense popul at i ons . Rich hunting grounds , favored fi shi ng sites , and the like , are always apt to be objects of content ion between even simpler societ ie s , but for agricultural civil i zations the basis from which their cultural blessi ngs flow--fertile river valleys and other types of productive l ands , mines and supply routes--must be controlled at a ll co sts . And those who are defeated can l ook forward only to exploita- tion , slavery, serfdom , and perhaps cultural annihila - tion . Thus , war becomes more deadly and serious , its consequences greater ; there is more to f i ght about , more to gain or lose (1 960 : 329- 30) . This type is "true" warfare . The reader will re call my di s - cussion of such di s tinctions in the se c t ion dealing with definitions . Type 4 warfare i ncl udes ·world war , which "may be regarded as a consequence of the industrial revolution" (1 960 : 330) . Coalitions of nations fight each other with huge armies supported by well- organized civilian populations , for 81 long per·iods of time . The II causes " for these wars are still tho se of Type 3 food- producers , with the addition of urgent i ndustri a l imperatives to have a ccess to raw materials and markets for fi nished goods (1 960 : 330) . To these four categories, one coul d add two more : atomic warfare and one of i ts consequences , l imited warfare-- total war but only to a ce~ta in point , i. e ., the use of nuclear weapons . Newcomb notes that wars between societies of different technologica l levels may take place because they are differ- ent--the poor want what the rich have , and the rich look upon the poor as fair game , e . g ., I ndians of North America versus European settlers . Yet Newcomb concludes without comment t hat whether or not a "culture goes to war" and uses its technological capability in war , depends upon a very wide range of other variables , which I shall discuss in a moment . Finally, this typology has nothing to do with i ndivid- uals , since culturologists conceive of cultuFe as "a superor- gani c entity, obeying it s own l aws and moving in accordance wi t.rJ. its own principles " (Ne -i,,~comb 1960 : 332) . ·when it is realized that cultures do have their own i nde - pendent integrity , their own direction and force , and an existence apart from individuals, the distinction be t ween the reasons why i ndividuals fight in wars and the ca,uses of t he ir culture ' s wars comes clearly into focus .... I n fact , a na tion may be able to produce more strongly moti- vated fighti ng men if they are i gnorant of the real causes of conflict . How many American youths could be enlisted to fight the batt les of the oil companies , or for the markets of southeast As ia? How many , on the other hand , could be enlisted to prevent the rape of 82 , their mothers , or to "see t he world " ... or for other personal reasons? (1 960 : 332 ) How do we account for this incons i stency between the etic and the emic? "Ali culture s pr-esent themselves " as moral , good , and true to their members , but relationships between cultures "have always been governed by the law of the jungle , by sheer power ," because " compet itive conflict " is the basic cause of war (Newcomb 1960 : 332- 33) . The exercise of for ce ma ,j eure may l ead to intricate rationalizations of behavior . Yet the student who views culture superorganically i s reli eved from being upset by the fact that "peace- lovi ng" nat i ons are frequently at war , or that (as the journalis- tic world assures us) the man in the street , whether i t be Red Square or Main Stree t does not want war although i t i s thi s man who wi ll become his nation ' s s oldi er . The superorganicist reali zes that the wishes and hopes of the indi vidual are the result of the interplay of cultural forces whi ch are affect i n him , and that i t is not the other way around Newcomb 1960 : 333 ; italics added . I n assessing the culturological contri bution to expl a - nations of warfare , the good points stand out , especiall y when considered in the cont ext of the intellectual history of anthropology . The emphasis on technology and its effects upon other cultural phenomena are not to be disputed ; nor the perhaps pivotal position in h~~a n life of materialism- -e co- nomics may indeed make the world go 1 round- -and the emphas i s on the consequences of the agricultural revolution ; nor the drive for a bird ' s - eye view of history ; nor the invention of heuristics to make that view communi cable and applicable . The culturological theory is part i ally true--these theor ists are not fools- -but false in ge nerality, as I will try to show. 83 My pri mary obje ction i s the loss s ome1here along the way of the "as i f " in the conception of culture havi ng an existence of its own , -i ndependent of human beings . Even though White , Newcomb , a nd , l ater on , Sahlins and Harris repeatedly state that culture cannot exist without man and that where man is , one will also find cul ture , they continue to ask a question that, I contend , i s unanswerable . As the biologica l de te rmini sts ask , "What i s man without culture like? , 11 the cultural determini s ts ask , 11What is culture wi t h- out man like? , " investigating the flip s i de of the nature - culture disjunction . No longe r do s ociocultural phenomena occur "as if 11 they were doing s o in accord with certa in pri n- ci.ples a nd l aws , but t hey rea lly do . '.I.'he mode l becomes the thing modeled : a metaphor of the human puppet responding to the tugs of the superorganic puppeteer ceases to be a meta- phor . Bidney ( 1953 ) discusses the problem of the culturo- logical 11 as i f 11 in detail , and concludes that the culturo:- logical fa llacy i ncorpo r ates a new animism of the superor- ganic , whi ch has expl anatory power as do all myths--and even perhaps a kernel of truth , as most myths--bu.t can hardly be called scientific , a na l ogous to theoretical physics . Pe rhaps one should qualify this by saying tha t it can hardly be ca lled sci entific , ~ - Poppe r states that his- toricall y nearly a ll sci entific the ories origi nated from or were ant icipated by myths . 5 Therefore , a theory found to be non- scient i f ic , i . e . , unt estable , may s till be important , 84 nbut it cann ot cla i m t o be backed by empi ri ca l evi dence i n the sc i ent if ic sense - - although i t may easily be , i n some geneti c sense , the ' re:sult of obse rvat i on '" ( 1968 : 38 ). Popper i dent i f i es Marx ' s theory of hi story as non- s cient i f ic or pseudo- scient i fi c be cause i t does not t ake predictive ,L risks , i . e . , be t esteQ f or s i tua tions i n whi ch i t may not apply . Poppe r a ccuses Mar xi sts of a dj usting the t heory s o t hat i t will a l ways be demonstrat ed , like t he ast rologer or soothsayer maki ng open- ended , adjustabl e pr e dictions s o he wi ll no t lo se clients by b eing proven wrong . Since Whi t e has been s o strongly influenced by Mar xi s t the or y , Popper ' s crit icisms a l s o appl y , not a nal ogous l y , but l iter a l ly t o hi m. The c r iticisms a r e especially relevant when one recalls that Whi te ' s j ustificat i on f or "as i f " statements i n t he f i rs t pl a ce i s i n t he name of s cient i f ic expl anat i on and a "s ci- ence '' of culture . At tempt i ng to be what he conceives of as sc i ent i f i c, he create s myt hs i ns t ead . I n pa r aphras i ng the cul turo l ogica l view of war , then , when men j oin t oge ther t hrough choi ce or cons cr i pt i on , ri de or march off to ambush or engage a n enemy , carrying spears o.r M16 1 s , and give as reas ons for doing so right i ng a wrong , gett i ng r ich , gai ning respe ct and pr est i ge , doing a j ob , or f oll owing an order , their act i ons , beliefs , and materie l are s i mply expressions of cu l tural l aws , a s f a l ling rocks and feathers express the Law of Gravitat ion . 85 Ordinarily , one does no t expect l aws to explain why caltural systems behave in such a manner , or why thi ngs fall- - b-c_t to simply s t ate that they do so , regularly, and that this regularity i s testable . But testable under what conditions? The l aw of fre ely falling bodies operates i ri a vacuum ; the di sruptive factor of atmospheric context is removed . Rocks dropped from the Tower of Pi sa or the Empire State Building will reach the ground before feathers ; a coin •will fall almost as fast as a cannonball , but in a vacuum all will fall at exactly the same rate . We can drop feathers , r ocks , chairs , and flowers thousands of times and they will behave as the l aw states . If they did not , the law woul d have to be altered or thrown out, its new form ub ject to similar ly risky tests . 6 Let us set up these " l aws ," one physical , one cul- tural , side by side . We should be able to do this if , as Manners and Kaplan contend , there is no "logical or ontologi- cal gap" between the physical and the social sciences (1968 : 10) . For the f irst case, I shall continue examining the law of free ly falling bodies for several reasons . White uses it as a standard of explanatory power ; it is relatively simple ; it is i mportant in the foundation of classical me chanics and classical physics ; and it i s determinist ic--characterj_sti cs making it more analogous than say quantum physics to attempts at macro-theory construct ion i n anthropol ogy . The law states t hat the velocity of a falling body is proportional 86 to the time of it s fa l l and that t he di stance covered i ncreases as the square of t i me , presented i n the f ormula : s =½a t 2 where ~ equals the distance covered , 1 equal s t i me , and a e quals a cce l erat ion , a constant . ( The a cceleration of fre e f a ll i s 386 . 2 i nches per se c ond squared , or 386 . 2 i nches per s econd per s econd , varyi ng sli ght l y wi th l a t itude a nd a lt i - tude (Gamow 1962 : 33 ) . ) Simple , predictive , mathemat i cal , and t he bas i s for Newt on ' s Law of Universal Gravi t y7 a nd Ei nste i n ' s the ory of gravi tation as t he curvature of t he space- t i me cont i nuum. There i s no need for an "as i f " qualif icat i on ; such a qualif icat ion would be i n error . For the s ec ond case , consider Kapl an ' s Law of Cultural Domi nance , which a t f i rst l ooks like a l aw , i . e ., it s t at e s a regul ar mode of a ct i on of a class of phenomena , but i t i s a di rect homolog t o the Darwini an the ory in which t he pri nciple of natur a l se l ection i s no t a l aw . Kapl an ' s "l aw" states t hat the ... cul tQral s ystem whi ch more e ffe ct i vely exploits t he ener gy res ources of a g i ven envi ronment will tend to spread i n that envi ronment as the expense of l ess effe c- tive systems .... a cultural system will tend to be found pre cise l y i n those envi ronments i n whi ch i t yields a higher energy re t urn per unit of human l abor t han any a l ternative s ystem avai l abl e (Kaplan 1960 : 75-76 ) . Then Kaplan proceeds to explain that he real izes that as ye t anthropology has not devi sed a measure of thermodynami c effect i veness , ot her than the extent t o which a cult ure i n a 87 particular environment dominates or is domi nated . In the present state of evolutionary theory we are pl a ced in somewhat the same embarrassing pos i t i on as t he bio lo- gists who account for the survival of certain organi sms in terms of their better adaptability and then turn about and assert that the reason they know one organi sm i s better adapted t o i ts envi ronment than anothe r i s t hat one survi ves and the other does not (Kapl an 1960 : 76 ) . Oue shoul d certai nl y not think that such effibarrassrnent i s an i nd i cat ion t hat adaptat ion i s a wor t hl ess con cept , becaus e i t i s not . I ts uti l ity has been demons t rated repeatedly . I ndeed , we woul d not be abl e t o talk about t he processes of biologica l evolution wi thout i t . But the i mportant poi nt i s that Kapl an apparent l y c alls thi s t rend , or proposi t i on , or even pr i nctpl e or theory a l aw be cause of t he i mplicit assumpt i on that whi le it cannot qual ify f or l aw stat us now, i n the future we wi l l be able to fi ll in t he hol es . On t he bas is of i ts potenti al , ca l l i t a l aw . Us i ng such reas oni ng , plus an uncondit i onal c r i terion of regul ari t y , we c oul d ele- vate any nwnber of s t a t ements t o the s t atus of law and thus by f i at be scientific . Why not have t he Law of Oedi pus , t he Law of Frustra tion and Aggression , the Law of Maximization , the Surplus Law , the Law of Male Dominance, and t he Law of Matri l ateral Cross-Cousin Marri age ? Harris (1 97 1 : 203 f f . ) does offer a mathemati ca l meas- ure of te chnoenvironmental efficiency , which coul d be c ons i d- ered a direct measure of adapt i on . Hi s formul a states : E = ill X t X r X e 88 where E e q_uals food energy or the number of c a lo r ie s produced per year ; rg equal s the number of food producers ; .1 equals hours of work f or each- foo d producer ; ~ equals ca lories expended pe r hour ; and Q equals the average number of ca lories produced fo r each calorie expended . The value of e i s derived by : total ca lories of foo d pr oduced per day the number of workers x average hours of work per worker x 150 calories ( that which each worker expends each hour above basal me t abolism) The refore , 7.4 Kung Bushmen of the Kalahari Desert working 6 hours each and expendi ng 150 ca lorie s each per hour , colle c t 64 , 200 ca lor ie s worth of foo d , g iving them an index of 9 . 6 . If t he i ndex were l ess than 1 . 0, one would assume that the people were starving . I ncludi ng this i ndex in the l arger formula , one is no t solving fo r E . I nstead , Eis estimated a t 365 x 64 , 200 (assuming of course that this i s a dail y average ) equals 23 , 433 , 000 . The value of~ i s given ~s 20 , t he number of adult i nd i v i dual s who worked a t food gett ing over a certai n perio d . The formul a i s completed by sol ving fo r !, whi ch is 605 hours per food colle ctor per year . One hundred forty- s i x Tsembaga Maring , swi dden agriculturalists i n New Guinea, produce 130 million ca lor i es i n vegetable foods annually , each worki ng 380 hours per year , and havi ng an effici ency ratio estimated at 18 . 0, and 18 million ca lo- ries i n ani mal food (pigs) , but whi ch produces a much lowe r efficiency rating of 2 .1 ( 1971 : 210-1 3 ) . Five mi l lion farm 89 workers i n the United States i n 1964 produced 260 trillion calories , each worki ng an average of 1714 hours , wi th an efficiency index of 219 . 0 (1971 : 217 ) . Harris spe cif i cally omits energy expended in food preparat i on (1971 : 206) . By interpolation with Kaplan rs Law , we could expe ct that the society wi th the highest efficiency rating i n a par- ticul ar environment wi ll tend to spread i n that environment and i n most cases by force or the. t hreat of f orce : military conquest, extermination , di spossession , coloni zation : ... an advanced cultural system can marshall a greater and more powerfull y equipped mili tary force,8 enabling i t to take , and hol d against encroachment or revolt , areas where its exploitative techniques are more effective t han rival systems (Kaplan 1960 : 88 ; note added) . Or a dominant type may spread by adoption , as a t hreatened society adapts to the system of the threat rather than fight- i ng it . Harris a~mits to the problems of i nexact itude i n wt.at one could call cynica lly a nutri t ional t heory of his- tory : Although this formul a i s constructed from several 11 guess- timates, " correspondences wi th t he data from other soc i - eties increase our confidence in it s basic a ccuracy . The most problematical factor i s the value of 150 ca lor ies per hour for~ - I t is very d i ff icult to measure ca lorie expenditure per time uni t under natural f ield c onditions (Harris 1971 : 205) . One would also have to know t he caloric values of nat i ve food s , the caloric i ntake of each worker , the exact amount of time spent working , and seasonal variations . Unlike taking a pulse by count i ng for 15 seconds and multiplying by 4 , i n measuring highly vari able work , as opposed t o a position on 90 an as se mb l y line , it does not se em legitimate to me a sure for 3 mouths a nd multi ply by 4 f or an annua l figure . One should know these things beyond "guesstimates " if the mathemat ica l rigor of the f ood-energy formula is to even approach that of the Law of Falling Bodies . Wha t might be empirical i mplemen- tation of the Law of Cultural Domi nance outwardly wears the trappings of physical law but upon closer inspection i t is fille d with unknowns . There are 2 possible courses to take i n deali ng wi th these unknowns . The first i s to invent means , expend effort , and demand rigor to fill in those unknowns as Koebben (1 967) , Hempel (1 959 ), and Harris insist on the assumption that the task is possible. The second i s to examine t he possibility that at the most such formulae are estimates and will remain so because of the nature of huma n life, i n this case "work ." Putting i t ano t her way the unknowns can never be filled in exactly, like the rate of acceleration on earth , because human action is open er..ded , not because we have not yet found the limits . For exampl e , what about time a nd caloric expenditures for events like thinking and talki ng about food- getting , and pre parations for food-getting activitie s ? We cannot, con- tra ry to White ' s belief , put human action in a vacuum, nor is it legitimate to operate "as if" we could . Many an earnest psychologist or sociologist apologizes for t he fac t that his science is not , or is not yet , a science i n the Newtonian sense, that i s , a s mall set of formulas and principles from which individual behavior can be predicted in much t he same way t hat astronomers 91 predict an eclipse . But are apo l ogies ca lled for? In Newton , as i n Einstein later, t he fullness of time brought together a genius for theorizing anct a domain in which certain kinds of cyclica l or otherwi se repeatable events lend themselve s to mathemat ica l treatment . How- ever , not ever . sc i ent i f ic sub ,ject exhibits that i nitial kind of order Niels en 1967 : 34 ; i tali cs added ). Yet Service (1 960 ) l auds the culture evolutionary point of v i ew as good becaus e i t i s pre dictive and therefore relevant to modern life i n forecasting the future and making i mprove- ments i n the world . Agai n , usefu l engineeri ng i s the ultima r a tio . Before exami ning the pr edict i ve and expl anat ory c l a ims of the Law of Cul ture Dominance-- The l aw of cultural domi nance , which i s derived from examination of the pro cess of the r i se and spread of dom- i nant culture types , not only underlies the distri but ion of culture and the hi s t oric movements of peoples and societ ies , but also explains why some cultural syst ems have been able to s read a t the ex ense of others and some have not Kaplan 1960 : 92 ; italics added --I want t o look at another comparative case , from the socia l sciences this time . Formal economi cs , commonly thought of as the mos t "scient i f ic " of the social science s , makes a strong engi neer- i ng claim, but when looked at closely i ts i deal postul a tes operating in a social vacuum do not explai n or predict well i n the real worl d . We mi ght even ca ll f ormal economic theory " super- etic ." The world as depi cted by convent ional e conomi cs is a highly "idealized " world . I t is a worl d in which indi- vidua l s act with complete information and foresight ; i n whi ch all action issues from economically rat i onal de cisions and is dire cted toward ends . that are always maximized ; in which there are no cultural or psychologi- ca l restraints on translating de ci sion i nto imme di ate 92 action ; and in which all i r~di victuals make choices and act wholly i ndependent ly of one another . Within this ideal- ize d worl d , economists have been able to move with logi- cal consistency , deductive certainty and , frequently , mathemat ica l elegance . I n respor~ding to criticisms that this i dealized _wor-1d seems to bear little relat i onship to any concrete empirical system , economists have replied that thi s is the way of science (Kaplan 1968 : 237) . I n his elegant criticism of this view , Kaplan points out that economic rationality is not a limiting condition of human behavior in the s ame sense that a vacuum is a limiting condi- tion of physical behavior . The buyer s , sell ers , consumers , · and entrepreneurs of microeconomic t heory are not real people but idealizations and abstractions , assigned certain proper- ties ·within the theory . Even if one does not reject formal t he ory a priori and plugs substantive data into it , it only explains and predicts i n market economics . Yet there is serious question as to its usefulness i n market economics , spe cifica lly the inability to move from ideal cases to real ones . So long as economists have remained in t he i r purely formal-hypothetical world they have been abl e to explain and predict with some measure of success . When they have tried , however, to make the transition to any concrete economic state of affairs they have encountered the same methodological problems that other social scientists con- front (i. e . , non-closed systems , a mult i tude of vari- ables , etc . ) . Their predictive successes seem not to have been appre ciably greater than those of othe r social scientists .... (Kaplan 1968 : 241) But 1£ evolutionary theory, biological or cultural , predictive? Recall that Nielsen has stated Darwi n ' s theory to be nonpredictive , i . e ., it ca nnot tell when or i n what direction a species will evolve or under what spe cific 93 condit i ons , or even that evolution will cont i nue to occur . It does explai n what has taken pl a ce i n the past and , i f change occurs i r~ the future , how it came about . Yet t he t heory even expl a i ns only i n a general wa y . Darwi n ' s propos i t i ons do not directly expl a i n any of the facts tha t prompted him to f ormulate them. . . . I nstead , a conne c t i on between the theory and a particula r fa ct is made by talking through to the fact in a manner sugge s ted and pe rmit ted by the theory. This talking (or writing) is not to be confused with any form of [mathematica l] ca lculating .... (Nielsen 1967 : 33- 36 ) For example , in looking at a collection of fo ssil hominid s kulls , we notice cons i derabl e variation i n dentition and f a cial s kelet on . I ndeed , through seriation, supported by a bso lute and r e lative chronologica l evidence, the spe cimens are arranged i n what i s thought to be a developmental se~uence . The observed vari ations through time are e xplained i n evolutionary terms as due to adaptat i on to and se lection for both vege table and meat eating . But we cannot deduce t hi s explanat ion i n t i ght logica l form from the principle of natural se lect ion which states that t hose i ndividual organ- i sms best able to ge t along in a particular ,environment will tend to live l onger and produce more offspr i ng than those who are less well able . I nstead , we work from the t heory to t he fact in what Nielsen calls a "di s cursive l anguage " using our knowledge of physiology , genet ics , and culture, e . g . , the r e l at i onship between the development of the hand and tools and both the re duction of big gri nd i ng mol ars and big canine s . Predictions about future homini d dentition are not i mpo ssible, 94 but they cannot be deduced from the theory . For instance , similar to Darwin ' s predict ion about the extinction of tame foxes i n the fac e of human intrusion , in appraisi ng Western middle-class dietary and dental hygiene habits plus increas- i ng longevity , we can predict a continuat i on of peri dontal disease and tooth loss despite our enormous dental care i ndustry . We cannot predi ct , however , what hominid dent ition will be like 10 generations from now . I n an art icle anal yzing the expans ion of the Tiv and the Nuer against their respective ne i ghbors , Sahlins (1 96 1 ) uses an etic grid over anemic case with results similar to but more sophi st icated than Newcomb rs analysis of Plains war- fare . Sahlins argues that the crucia l factor i n the expan- sionary success of these 2 soci eties i s not sufficiently their environment , technology , and economics but the ways in ~~ich people are organized i n social groups in order to adapt to certain c ondit ions . The Tiv and Nuer are both expanding against popul a - tions whose subsistence base i s the same as the i r own , but while the Tiv are the l arger popul at i on (800 , 000 , the l argest group i n northern Nigeria ), 200 , 000 Nuer have intruded suc- cessfully i nto the pasture l ands of 900 ,000 Dinka . The sig- nificant factor apparent ly i s the segmentary lineage system by means of which people are able to mobilize and through which they make claim to the right of l and use . Among the Tiv , " .. . every compound headman within the mini mal segment 95 holds a right against the world to sufficient farming l and " (Sahlins 196 1 :337)c The worl d cons i sts preferabl y of for- eigners , ~ut for those lineages towards the cente r of Tiv- l and , one moves against those ne ighboring lineages most di s tantly r e l a ted to one ' s own . As internal lineages grow in population and sat i s fy their f elt right to land by taking l and from distantly related lineages , eventually those kin groups at the borders are forced to replace their lands expropri ated by other Tiv by moving against foreigners . Such act ion l ea ds to a "long and bitter war" (Bohannon 1954 : 7 ). The residents of the borde r villages may be moving a gainst t heir wj_lJ_. "The lineage is simply crowded out as the Tiv side of its land i s consumed by the appetites of other Tiv" ( Bohannon 1954 : 7; italics added) . The Nuer , whom Sahlins cons i ders as "perhaps an out - standj_ng i nstance of the Law of Cultural Dominance," appar- ent ly i ntruded into land held by the Dinka and successfully pushed back , split up, and absorbed many of the autonomous subtribes characte r istic of Dinka so cia l organization . As i n the Tiv case this succe ss i s due , Sahlins a rgues , to Nuer segmontary lineage organi zat ion. The Dinka , even with a pop- ulation 4 . 5 times as great , are unable to mobilize agains t continued Nuer expans ion because of t he absence of segmentary lineages , and the Dinka , even i n the face of grea t pressure , have apparent ly not invented an a lternative . In Sahlins ' words , 96 The DiLka lack the t uermostatic mechani sm for massing against the outside , a deficiency that has been fa a l ( 1 961 : 340) . The reason for this difference between 2 cultures that are otherwi se alike i s apparently be cause the Dinka arri ved i n t he region first and , fa cing no opposition , dispersed i nto small autonomous groups wi th little need for cooperat i on [ quant itat i ve utili zat i on of negat i ve entropy]. The se ci r cumstances favor fiss i on but sele c t against com- pl ementary opposit i on or fus i on , and long t erm occupat ion will eventuall r fix t his stru ct ure maki n i t c om arativel i nf l exible Sahlins 196 1 : 342 ; ital ics added . The Nuer , however , were i ntruders into an a l ready oc cupi ed area and fa ced opposit i on . "Thi s se l ective cir cumst ance pl aced a premium on the ability to fuse as well as t o seg- ment , on complementary opposition" (Sahli ns 196 1 : 340 )0 I f the Nuer deve l oped segmentary lineages as an "adapt i ve response " to their intruding , i s it not reas onable t o ask why the Dinka did not deve l op an adaptive response t o i ntrusi on? On the one hand , we ca n accept the proposi tion t hat infl exi - bi l i ty be cause of age means they coul d not and t herefore were pushed around and threatene d with ext i nct i on--t he y we r e t oo specialized . On the other hand , one could interpret Dinka capitulat i on and abso r pt i on as an adaptation , i . e ., an alternative to being kill ed , but be subjec t to the adaptation tautology . Why did the Tiv and Nuer move agai nst their ne i ghbors at all? Sahl.ins cites overpopulation as the reason given by 97 both Evans- Pritchard and the Nuer themselves . Yet amo ng the Tiv, while part of their land i s overused i n the sout h , rapid expansion has recently occurred in the north , where popu- lation density i s less than half the average of 64 pe r- sons per square mile. The s i gnificant factor seems to be tne i dea and beli ef of the people themse lves that they need l and . Sahlins concludes : it seems to us that a certain relativity i s required i n assessing land hunger among societ i es compet ing for occupation of a spe cifi c habitat . Because the success of one contestant is ne cessarily to the detriment of the other , neither has enough land until the other has been eliminat ed . The need for "living- space" i s built in : it be comes a cultural att itude and theory , particularly in that society which has the dec i sive compet itive advan- tage . Among the i nvaders a natural 'i ncrease of popula- t i on beyond the carrying capacity · of present resources will be taken for granted , and at l east for them land hunger exi sts--tr~_i dea i s adaptively advantageous--even if, by objective standards , there i s enough l and to sup- port the present populatioL ...• From an adaptive point of view this is no paradox (Sahlins 196 1 : 341). Sahlins 1 explanation does not seem to follow, especially the statement "Be cause the success of one contestant is neces- sarily to the detriment of the other , neitber has enough l and until the other has been eli rn i nated , 11 and the s t atement , "From an adaptive po i nt of view there is no paradox ." The former i f ... then statement does no t logically follow , even t hough taken separately t he pr emi se and conclus ion may be true ; t he l a tter adaptive statement is by nature non- paradoxical and unfals ifiable. On the one hand the Tiv are not maki ng a real adaptive re sponse because there is p l enty of land ; on the other hand , 98 be cause they think they need living room and more l and , moving agai ns t others i s an adapt i ve response t o Tiv and Nuer i deas of native concept i ons of overpopul at ion and land need . One i s l eft with a " so what " feeling i n regard to the explan- atory power of a daptat ion but with qui te a different reaction to the explanatory power of the reasons people give for doing certai n things . With some hesitation Sahl ins does consider such information significant--and Harris does not , as we shall see--and that ideas can be "adaptively advantageous ," i n the sense here that people are mot iva ted to continua lly i ncrease their land holdings . Harris wants to supply a ma cro- theory of soci ocul tural evolution , the basic principle of which is not a l aw like Tewton 1 s Laws of Moti on or the l aws of quantum mechanics, but a "law " analogous to the principl e of natura l selection in Darwinian evolutionary theory , " . .. a basic re search strat- egy , f rom the application of wi1ich there is an · expectation that a nomothetic causal understanding of sociocultural phe- nomena may be achieved ." Harris 1 s ociocultura l analog to the principle of natural se lect i on is the pri nciple of te chno- e nvironmental and techno-economic de termi n i sm . This principle hol ds that simi l ar technologies applied to similar environments tend to produce s i mil ar arrangements of l abor i n production and distri but ion , and that these i n turn call forth similar kinds of social groups , which justify and coordinate their a ctivit ies by means of simi- l ar systems of values and beliefs (1 968 :4) . 99 At this po i nt one could utt er Louch 1 s "so what t s new" re spons e . But there is more to it . As differential repro- duct io n i s to natural selection , cultural materialism is to techno-environmental and techno- economic determinism . Based on Ma rx 1 s The Critique of ~olitica l Economy , This strategy [or "law 11 of cultural evolution] states that the explanation for cultural differences and s imi- larities i s to be found in the techno- economic processes respons i ble for the production of t he material require- ments of biosocial survival ... t hat the techno-economic parameters of sociocultural systems exert selective pres- sures in favor of certain types of organizational struc- tures and upon the survival and spread of definite types of i deologi cal complexes ... that in principle , a l l of the ma jor problems of socio cultural differences and simi- larities can be solved by i dentifying the precise nature · of the se l ect ive paiameters ; yet as a general principle , it does not commit itself to the explanation of any spe- cific sociocultural types or any speci fic se t of i nst itu- tions (Ha rris 1969 : 241) . If It the explanat ion of biological transformations is to be found in the adaptative advantages (measured in terms of repr oductive success) , which part i cula r innovations9 con- fer upon the organism and its lineage " (Harris 1968 : 24 1) , then the explanation of cultural transformations i s to be found in the adaptative advantages (measured in terms of thermodynamic efficiency ) which particular inventions confer upon a society [ the Law of Cultura l Dominance] . Harri s then proceeds to offer a materialist explanation of the difference between race relations in Brazil and t he United States , as opposed to an i dealist one based on Portuguese national char- acter and Catholicism versus Anglo- Saxon racism and Protest- ant i sm , that i s based first on ecol ogy, t hen on migratory 100 pat tern s from Europe (la bor ), demography , industrializat i on , polit i cs , militari sm , and cognition . Perhaps he wou l d con- s i der explaini ng nat ional character a nd Angl o-Saxon racism in techno-envi ronmental te rms . I doubt that a nyone woul d a rgue that an explanation based on national character or raci sm cou.ld a ccount for t he historical data nearly as well as Harris ' "ma t erial " explanation . Nor do I expe ct denial of t he great i mportance of ec ology and economics in human his tory . But what are the l ogical entailments of using cultura l materi a lism as a unitary t heory or t he true explanat i on? One t hi ng one mus t not do i s pay attenti on to the reasons people give you for t heir a ctions--that i s emic and does not tell us what people actually do . Cultural materialism , Durkhe imian sociology , British s ocia l anthropology , Freudian psychologi- cal anthropology , and Fre nch structurali sm- - especially that of Levi- Strauss-- ... are predicated upon the a ssumpt ion t hat t he actual parti cipants in soci al l ife a re i ncapable of an objective de s cr iption of their own behavior or of a scientificall y valid explanation of that behavior . All of these approa ches t hus share a common COF.ill1itment to cleari ng away the errors of autoanalysis , the facade of i deology , t he rationalized appearances of things , in order to pene- trate i nt o t he deeper l eve l s of both thought and a ction . The announce d goal i s to explain social facts in terms of social facts , rather than i deas in terms of ot her ideas .. . . The hypot hes i s t ha t causal expl anat ions reside in the material condit ion of life en joins an attitude of extreme skepticism toward the relevance of the manifes t meanings of a ll verbal events (Harris 1968 : 234) . 101 The consequences of this asswnption are enormous and far- reachi ng . Surely, if Marx , Engels , and Harri s asked a Cheyenne wny he spent so much time fletchi ng an arrow with a ce rtai n kind , amount , and positioning of t he feathers , they would not doubt hi s r epl y that not to do so would mean the arrow woul d not fly straight , he could not hunt game as well, and hi s f amily might go hungry . But what i f they asked hi m why he a nd his fellow Cheyenne perform an annual ce remony to r enew the Sacred Arr ow bundl e? Woul d they doubt his state- ment that no t t o do so would put the whole tribe in spiritual jeopardy , so their arrows will not fly true and t he i r fami- l ies will go hungry? Probably . They mi ght reply that the rea l reason is that Cheyenne t e chno l ogy (arrows ) and economy (hunt i ng ) have exerted "sele ctive pres sures ... upon the sur- vival and spread of definit e types of i deologi ca l c omplexes " (Harri s 1968 : 241 ). How does one determi ne whi ch s t atements to doubt and which not to doubt ? If one subscri bes to t he hypothesi s "that causal expl anations reside i n the materi a l conditions of l ife ," then a materi a l ist reply given by an informant i s accepted and others r e j e cted as being superficia l i diosyn- crati c r at i onalizations . But he may be wrong about his material ist expl anation ; a ft er all , he i s deluded about why he performs t he religious ceremony . We kno1v he i s de l uded because t he i deas or values of religion and spiritual jeopardy and sacred arrows do not expl a i n t he facts of 102 success or fa ilure i n hunting-- i n the materialist paradigm . At t hi s point we are entering t he real m of yet anothe r di s - junc t ion , fact - value , an especiall y i mportant i ssue i n the social sciences . While I cannot take up the issue here for the attention it deserves, le t me note that anthropol ogists of whatever t heoretica l persuasi on by a nd l a r ge are deeply attached to this disjunction and accept and teach i t as enlight ened doctri ne , which in the face of absolutism I suppose i t is . On the specific is sue of warfare , Harr is offers an expl anation that appears to have 2 i n t erre l ated components : (1 ) compet ition for resource s to support growing populations ; ( 2 ) an adaptation to maintain populations below " techno- environrnental carrying capacity , " the absence of which would mean populat ion contro l through malnutrition a nd disease . Both of these component s are be s t understood in evolutionary terms , specifica l ly principles analogous to that of na tural sele ction--discussed to some extent above . The first compo- nent i s straightforward and cons istent with t he economi c explanations di s cussed below. The second is more difficult a nd it s implicat ions less clear . Harris sees warfare as the i nevitable result of com- petition for natural resources t o support food production , which in turn l eads to increased populat ion , i .~ ., the enlarged foo d base precedes a rise in populat ion . Subse - quent l y , smaller a nd therefore weaker poptilations , in orde r 103 not to be eliminated by l arge r stronger ones, must i ncrease their own numbers , which leads to i ntensified friction between populat ions and eventua lly to warfare . Food- collectors , be cause of their SL,a ll dispersed groups with fluctuating membership and network of marriage alliances, have conflicts but not "true" warfare . Food-producers , how- ever , experience intensifie d warfare , in competing for crop l ands . In- group identities increase and as a result whole vil- lage communities become each other rs "enemies. " The fre - quent occurrence of full- scale wars of annihilation by one low-energy a gricultural community against another cannot be denied . These wars differ from modern wars only in scale , in effectiveness of the weapons of homi- cide , in degree of the organi zation of military exploit s . If primi tive wars were less genocidal or less brutal than our own , it was simply for lack of technique and technology (Ha rris 1971 : 226-27 ) . He assmnes that the expansion of "high-energy " societies ( those with higher productivity) is determined , and that preindustri a l ... sociocultural systems are under constant ideological and political pressure to expand to the limit of their technoenvironmental carr;ying capacity . The l arger the group , the more se cure i t is against attack . Hence in a g iven r egion each group tends to maximize its s trength by getting as close as possible to its loca l carrying capacity (Harris 1971 : 225) . These assumptions , that expansion and attack are inevitable , are quest ionable : some sociocultura l systems do , some do not. Furthermore , I am not clea r whether or not maximi zing population is supposed to be "unconscious " or conscious . There are exampl es in modern history where large f amilies and 104 ideologies of "fi l ling up empty space " are consciously pl a nne d a nd carried out , e . g ., the settlement of frontiers by Euro- Americans , Germany in the 20th century , and Russia aft er World War II, but I am not clear on comparative cases for non- We s tern societies . The determination of "carryi ng capacity " itself is a diff i cult one , with probl ems similar to those of determini ng te chnoenvi r onmental effici ency and economic surplus . Harris ' se cond component i s l ess clear , be cause he vi ews warf are as an e cological adaptation of popul ation con- trol analogous to natural sele ction for bipedalism and the capacity t o make t oo ls i n biol ogi cal evol ut ion , int erl a ced wi th cons cious human motives such as contracept i on . "Si nce primitive peoples lack effective chemi cal or mechani cal con- traceptives , all primitive systems of populat i on control alternat i ve to war also i nvolve sufferi ng , deprivati on , and the reduction of human we l fare " (Harris 197 1 : 229 ) ~ Mal nutri- tton and di sease as a "system " of populat i on c ont r ol .. . [ are] e c ol ogi cally even more dangerous than primitive warfare s i nce the ent i re environment may be come degraded through overuse , resulting in a permanent and i rrevers- i ble declj_ne toward extinction . Another alternative t o war among primitives is to practi ce i nfanticide . To con- trol popul ation by ki l l ing one 1 s own children , however , can s carcely be regarded as a ma j or i mprovement over war - fare from the point of view of human we l l - being (Harri s , 197 1 : 229) . While contraception and infanticide are conscious means of controll ing population (e . g ., the mother who smothers a new- born be cause it will mean too- early weani ng and weakened 105 health for the previous child gives this as the reason for her heartbreaki ng a ct), reducing population is not given as t he rea son f or goi ng to war except i n rare i nstance s s u ch as genocide . Obviously , a group woul d not go to war to reduce it s own populat ion . Harris rea lizes that the public evidence for warfare as a population contro l mechanism is v i rtually nonexistent . In a general sense , we can attribute the underl yi ng causes of primi t ive warfare to popul a tion pressure in conformity with the t he ory [ given above] . I t must be admitted , however , that t he evidence support i ng this theor y has been obscure d because t he motives that the belligerent s themse lves cite for going to war rarely i ndicate any awareness of population pressure (Harris 197 1 : 227 ; italics added ) . Harris explains the lack of fi t betwe en the etic and emi c: t he reasons peopl e do give- - revenge for homicide , trespass- i :...-1g , poaching , witchcraft , adultery , and woman- stealing--are manife s tations of social and physical di st res s as the popu- lation saturates its technoe nvironrnental carrying capacity , and that t he acts provoking such revenge can "consciousl y or unconsciously ... express the need of a group for more territory" (Harris 197 1 : 227) . It may seem strange t hat the people who lose their lives i n armed combat se l dom a ccurate ly unde r stand why t hey do so . But the masking of deeper cause s by superficial psy- chologi ca l motives is advantageous for groups l ocked i nto a system of population control t hat depends on war . To understand the causes of war i s t o relieve the enemy of the onus of guilt , making it i mpossible to mobilize the adrenalin and other hormones ne cessary for effective hand- to- hand combat . The primitive group that i s bur- dened by doubt s i s subje ct to anni hila tio n (Ha r ris 1971: 227 ) . 106 Thus , Harris has tied up the loose e nds . Not only i s warfare adaptive , but not knowi ng the real reason why one goes to war i s also adaptive , otherwise one cou l d not fight and woLJ_ld be eliminate d . But re call in the discussion of Sahlins that the Tiv and Nuer go to war because tney think they need l and primarily , even though there may be no real need . I cannot conceive of indi vi dual human beings "locke d i nto a system of popul at i on control. that depends on war ." I can conceive of them i nfluenced by beliefs and values , varying with individ- uals, e specially the belief where one life and one wrong must b e pai. d f or wi. th ano the r 1 1· f e an d re t ri· b u t i· on . 1 O Finally, is it illuminating to treat the causal effects of war--reduct i on of popul at i on--as causes, which peopl e ' s reasons , purposes , and intentions merely "mask "? But Harris (1972) in l ater publicat ions has clarified his theory of primitive warfare and population control , largely because of the influence of research by Divale (1 970; 1971) , who in turn has based his theory on Chagnon ' s Yano- mamo work . I will discuss Diva l e ' s findings i ~ a moment , but first a restatement of Harris ' theory of primitive warfare as a population control device . Whe reas in 1971, Harris rejecte d infanticide as the primary means of population control , i n 1972 he has changed his expl anation so that female i nfanticide is the initial response to population pressure , and warfare is a conse quence of this i nitial condition . Thus, when a given popul ation 107 begins to expe rience pressure on resources , i . e ., it is reaching te chnoenvironmental carryi ng capacity , i n t he absence of effective c ontraceptive and abortive technique s , i nfanticide , whe the r as a conscious act or t hrough neglect, is the easie s t way to control population . Simpl e neglect of babies is perhaps the most common form of population control . This will begin to t ake effect at a point well be low maximum carrying capacity as mothers , burdened by extra work , become l ess responsi ve to the demands of their children . The babies cry unattended for longer periods and t he mothers nurse them less effec- tively or le ss often . In ecological perspective , the line separat i ng i nfant negle ct from infant icide is extremely thin . In few primitive cultures will the mem- bers admit that the murder of children i s common . But unconscious deprivations can exert a s much i nfl uence on i nfant mortality as deliberate i nfanticide (Harris 1972 : 18) . Th i s proposition would appea r to be readily testable , if not with pri mitive soci eties , then with the l arge number of pop- ulations i n the thi rd world that have rea ched or already passed the carryi ng capacity of their envi r onrnent . 11 Perhaps such child-care studies already exi st ; I am not familiar with the literature . Research on s uch a problem should be carried out by a woman i n t he field who shoul d probably work solely wi th women i nformants--as evidence for trus tworthiness at t he very least : t he r e are many i mportant things that members of one sex simply do Lot want members of the opposite sex to know about. The s i gni ficant qua lification t o the practice of neglect or i nfanticide i s that it i s directed agains t femal e babies . Eti cally , Harri s expl a i ns thi s as an ecological 108 condi tion of limi ting broo d stock--boys do not get pregnant . Emica l ly , the de ci s ion to kill or the i nclination to neglect fe male rather than male babies i s an effect , i n feed- back fashio n , of the consequences of the i nitial case of fema l e i nfanticide . That i s , due to female i nfant icide there will be a shortage of brides for the male s of that age group when they want to marry, so they raid other people to capture women , an a ct of war . Whether out of revenge or the desi re to rep l a ce those lost women , t he offended group will retali- ate i n kind . Once the practice of warfare becomes e stab- lished , female infanticide will persist , but now due to pres- sure to produce mal e warriors . Thus , a woman- shortage will persist and , as a consequehce , so will war . But Harris frames this chain of events i n functionali st terms : Thus the primary function of pri mit i ve warfare is not to kill off "surplus " male s but to i nsure the continuat ion of high level s of fe mal e infant mortality . Much of what we nowadays regard as mal e chauvinism has its roots i n thi s s i tuat i on . One conclus ion that I draw from this i s that the whole compl ex of masculine aggressiveness is a by- product , not a cause of war (Harris 1972 : 18- 20) . With this I must t ake i ssue . Pri mi tive warfare doe s not "funct ion " t o "i nsure " anythi ng . I wou l d say i ns tead that the prima ry effect of primi tive warfare is no t a reduction of a surplus male population , but perpetuat i on (if i t ever existed. ) of female i nfant icide because of the need for males to fight . Mortality is not a ll f emale ; mal es die , too , but as adult s , in the pursuit of women , g l ory , wealth , or whatever . 109 Unpleasant as i t may be , it i s difficult to avoid the conclus ion that warfare began as part of an e cologica lly adapt ive system of population control. Death through combat strikes us as was teful ; yet for primi t i ve pe ople s the a l ternative to war f or balanci ng the adult sex rat ios was the expansi on of male infant mortality [ an uncons cion- abl e a ct in many societies] (Harris 1972 : 20 ; i talics added) . But what is the evidence for that i nitial condition of population pressure to which female infanticide i s the re sponse? The data upoL which Harri s bases the above expla- nat ion are census data colle cted by Divale for 112 primitive popul a tions to demonstrate the hypothesi s t hat : ... warfare plays a po s i t ive and beneficial role i n pri m- itive society and i n fact i s neces sary for culture s at these l evel s of compl exity. Pr i mi t i ve warfare i s part of a syndrome which a l s o includes femal e i nfanticide, polygyny , and marriage a lliances . The almos t uni versa l occurrence of this syndrome i n primi t i ve culture s plus its important ecological role has led me to conclude that the syndrome constitute s the basic structural framework or templ a te of pri mit i ve soci a l organi zat ion . The syndrome ts purpose is to control excess popula tion and to maint a in an equilibri um between a group t s popula tion a nd their available resources given t he ir level. of techno- economic ability t o exploit those res ources (Divale 1970 : 2 ). The s timulus for Diva l e ' s hypothes i s i s Chagnon rs study of Yanomamo warfare and his i nfe r ence s about the r elat ionship between femal e i nfanticide and warfare (1968a , 1968b ). Di vale I s own evidence i s i nferential, ·using demographic fig- ures for 11 2 groups . Of these 112, 91 per cent have sex rat ios i n which boys outnumber g i r l s in the "young gener- ation ," with an overall average of 146 boys : 100 gi rls . He concludes that s i nce s ex ratios a t birth are a l mos t equal , t his di s crepancy can be explained only by the presence of 11 0 female i cfanticide (1 970 : 3) . 12 In the "adul t generation" t he rat i os are equalized or tipped in favor of females as mal es di e in warfare . For the 11 2 groups , the average adult rat i o i s 109 : 100 (1 970 : 3) , within what one coul d cons i der as a "normal" range . Unfortunate l y , the limits of that range are not specif i ed . How many more males than femal es s houl d one r equire before t he practi ce of infanticide can be i nf erred legi t i mat e l y , if it i s not ment i oned i n the ethnographi es? While I do no t deny the existence or potent i a l s i gni f icance of fema l e i nfant icide , I woul d like to see some cons i derat i on taken of alternative expl anat i ons of unbal anced sex rat i os . For i nstance , among Orthodox J ews there a r e many more boys born than girl s . Re cent research on conception and sex determinab_on have expl ained t hi s parti cul ar case : f or the Orthodox sexual i nterc ourse is forb i dden until 14 days a f t e r the onset of a woman ' s menstrual period . I n a normal 28- day cycle , the 14th day marks the onset of ovul ation , and the usually a cid vagi nal s ecretions become a l kaline . Spermato z oa carryi ng a Y chromosome (which would determine mal e sex ) are l i ght er in we i ght , faster moving , and more sensit i ve t o unhospitable ac i d conditions t han those beari ng an X chromo- some . Thus , i f i ntercourse takes place on that optimum 14th day , the embryo i s more l ikely to be male . I ndivi dual physi- ological i di osyncracies must also be taken into consider- ation , and when they a.re , t hat theoretical 1 :1 sex rat i o of l i ve births seems unusua1 . 13 1 1 1 Divale readily concedes that there are serious prob- lems with the demographic data , but of another sort . First , a large part of his census data was collected long after colonial governments suppressed nat ive warfare , s o there are more adult males alive than he assumes woul d be under aborig- ina l conditions . Secondly , t he evidence for i nfanticide cannot be supported by ethnographic a ccounts but must be i nferred from the figures alone . After contact i n an unspe cified number of cases , ... the young generation ... sex ratios c ontinue to show high proportions of boys over girl s long after government control and prolonged missionary contact because infanti- cide i s much harder to detect and because most westerners --anthropologists i ncluded--do not l ook for i ts prac t ice . ... Primitive peoples quickly learn that government authorities and mis s i onaries frown upon and try to sup- press i nfantic i de so upon being questioned by even an anthropo logist they will of ten deny the practice. One such example are the Tsembaga of highl and New Guinea who denied using infanticide except against twins when ques- tioned by Rappaport ( 1968 : 15 ). However , their sex rat io for those under 15 years is 148 boys per 100 girls whi ch can only be explained by the use of female i nfanticide (Di vale 1970 : 5) . Undoubtedly , ttese would be diffi cult data to co l lect , par- ticularly if the anthropologists also happened to be male . Recall that infa~t icide i mmediate ly after birth i s carried out by women , not men . As for statistical data on male deaths in warfare , Divale cites 2 examples : 24 per cent for the Yanoruamo ( Chagnon 1968a : 140 ; 1968b : 20) , and 28 pe r cent for the Murngin (Warner 1930 : 481-82 ) . His best case comes from demo- graphic data on 5 Fj_ji I s l and tribes : i n 1880 the total 11 2 popul at i on of the tribes was 18 , 028 and the average childhood sex ratio was 130 : 100 . In 4 of the 5 tribes , the adult sex r at io was 96 :1 00 , but for t he fifth tribe, mal es continue d to outnumber female s 125 : 100 . Diva l e ' s explanation for the dif- fere nce i s that the fifth tribe had been w1der missionary control and without war for 40 years , while the other 4 tribes still carried on warfare (1 970 : 7 ; Fison and Howitt 1880 )174- 75) . Therefore , following Divale 1 s reasoning , wherever population figures indicate a ratio of more males t han females in childhood , against a base line of approxi- mate ly 105 males to 100 fe males , one can infer the existence of female infanticide; wherever the figures i ndicate a ratio of fewer males to fema les in adultho od , one can i nfer (per- haps unne cessarily fo r this variable) the prevalence of war- f are or , on the . contrary , its absence if there remain more males than female s o Up to this point, I grant that Divale ' s hypothesis i s attra ctive, but i ts testability has severe limitat ions , both in the na ture of the data and in Divale 1 s method , limitations that I shall continue to di s cuss in a moment . First , con- s i der the problem of percentages . Chagnon (1 968a : 140) col- lect ed data on the cau ses of 240 adult Yanomamo deaths . Of these , 39 , or 16 . 2 per cent, of a ll deaths we r e due to war and club fiehts , and of t hese , 33, or 23 . 9 per cent , were mal es . Co ntrast war fatalities with t ho~e of the biggest killers , malari a and epidemics (most ly malaria l ): 130 , or 11 3 54 . 2 per cent of the adult population , of whom 58 ( 42 per cent of all male deaths) were men and 72 (70 per cent of a ll female deaths ) were women . Only 3 out of 102 Yanomamo women died in childbirth . I suspect that this particular statisti c is deflated because of the complications that malaria can cause in a pregnant woman . Furthermore , Chagnon notes that his mortality figures are probably underestimates : The Yanomamo have very strong proscriptions on di scuss- ing the dead by name , and statistically adequate data requires complete genealogical information which is par- ticularly difficult to obtain with respect to individuals killed in warfare . Their anguish at the mention of killed kinsmen precludes i ntensive questioning on this topic (1968a :1 40) . One Yanomamo village of 200 people was raided about 25 times between November , 1964, and February , 1966 ; 10 people ( sex not spe cified) were killed, which is 5 per cent of the total populat ion . From a treacherous attack during a feast i n 1950 , 15 people or 13 per cent (agai n , sex not spe cifie d ) out of a population of 115 were ki lled (Chagnon 1968a : 14 1 ) . These f i gures indicate that primitive war i s often what it i s not supposed to be--lethal . Divale ' s stati stica l reasoning i n this vein may not be sound , however . I n an attempt at comparable stat i stical statements whi l e criticizing anthro- pologists for underestimating the deadliness of primitive warfare , he says : ... primitive populations are small and the male popu- lation is being controlled on a generational bas is . For example , it would be a large village that had 100 adult males but even so , i t would only take one killing a year over a 25 year period for one- fourth of those males to 114 perish in warfare . Thus the evidence indicated by the decline of males in the adult sex ratios pl us the specif- ically observed adult male death rates from warfare of about 25 percent leads to the co nclusion that primitive warfare was indeed effect i ve and functioned to regulate the excess male population (1970 : 8) . Following this reasoning, i f there were only 1 death per year over a 100-year period , 100 per cent of the adult males would have died in war . A death rate of 1 per cent per annum cannot be compounded into 25 per cent per 25 years and have the same meani ng ; the basis of the ratio i s dist orted . Before l eavi ng t he problem of the deadliness of primi- tive war , le t us compare Chagnon ' s Yanomamo figures to s ome from World War II, certai nl y a 11lethal 11 "modern" 11 war 1t by anyone 1 s criteria . Germany , with a popul at i on of 71 million, mobilized over 10 million soldiers , of whom 3 ,2 50 , 000 died ; that i s , 14 per cent of the populat i on was mobi lized , of ·which 31 per cent was killed outri ght or died of wounds . The U. S . S . R. , with a populati on of 175 million , mobilized 22 million , and lost 7 . 5 mi l lion soldiers plus 7 . 5 million civilians ; that i s, 13 per cent of her population was mobi - lized and of this 34 per cent died ; 8 . 6 per cent of her total population was killed or died in the wa r . Other horrendous f i gures include : Hungary lost 42 per cent of .her 350 , 000 troops ; Rumania lost 46 per cent of her 1 , 136 , 000 troops . For the Axis powers as a whole , 22 per cent of their 25 . 5 mi llion troops , r epresent i ng 11 per cent of a total popu- lation of 221 million , di ed . Out of t he total Axis 11 5 population , 3 . 4 per cent die d . For the Allie d or Uni ted Nations po\ ers , while Russia sustained the greatest losses , the United States wi th a to a l population of 135 million mobili zed 12 per cent and of t hose troops lost a mere 2 per cent . The United Kingdom , with a total popul a tion of 48 million , also mobilized 12 pe r cent , but l ost 9.4 per cent of t hem . Out of a total Allied population of 1 . 5 billion , only 5 per cent of t he population were mobilized and , of those t roops , 14 per cent were killed . Of the ent i re Allied popu- l ation , 2 . 9 per cent were kill ed or died i n war (Wright 1965 : 1 54 2) . Returni ng to the i ssue of i nferri ng female infant icide and sub equent warfare from sex rat i o data , I must fi rst point out that it is misleading to refer to the demographi c data frorr. 112 populations as a "sample ." Diva l e de c lares : The sample i s unbiased in the sense that it contai ns every age - sex r atio I was able to obtain in a year ' s search of the l i terature on pri mitive cul tures . Of the 11 2 societ i es , 91 per cent have sex ratios i n the young generation where boys outnumber girl s : the average for the 11 2 groups i s 146 boys per 100 girls (1 970 : 3) . Geographic stratification of t hi s sample i s as follows : North America , 34 s ocieties ; South Ameri ca , 6 s ociet i e s ; Africa , 5 societies ; Austra lia , 11 societies ; Melanesia , 56 societies . These societies would be better i dent i f i ed as populations and the who l e collection as a l most a caricature of the Tylor- Galton probl em . I .do not wi sh to disparage Divale ' s theoretical cont ribution to underst andi ng primit ive warfare , or to unde rvalue the i ndi vi dua l p i eces of 1 1 6 demographic data he offe s , but he cannot claim this study to be a quant ified test of hi s t heo r y . For example, Africa is represented so lely by 5 Nigerian popul ations : 1 represents Edo- speaking peoples , the other 4 are all villages of I bo- speaki ng people . As for Melanesia , 20 cases are from New Guinea; 13 are from the Bismarck Archipelago ; 2 are from the Admiralty Islands ; 5 are from t he Fiji I s lands ; 5 are from the Solomon Islands ; 1 is from Tikopia ; and 10 are from Great Andaman Island ! Of the 6 cases from South America , 2 belong to the Yumo of the Col umbia culture area, and the rema i ning 4 are from Amazonia--3 of the 4 are Cashinawa villages . Divale rs stat i stical inferences are limited to one : Of the 50 groups who were censussed [ sic] whi le i n an unacculturated state , 90 per cent have adult sex ratios of less than 11 0 males per 100 females . While on the other hand , of the troups who were censussed some time after their warfare period , 63% have adul t sex ratios of ove r 11 0 mal es per 100 females . A chi square test indicates over a 99 percent probability that the corre- lation between the adult sex ratios and the presence or absence of warfare is not due to chance alone . A Q test and a Phi te st a lso show positive associat i ons and corr e- lat i ons (1 970 : 4) . The co rre l at ions are X2 = 33 . 67 ; Q = . 879 ; phi= . 542 (1970 : 11) . Dival.e provides no table showi ng distributions of ce ll frequencie s , nor does he speci fy whi ch are the unacculturated and a cculturated societies or what c onstitute s "young " or "adult . " Because of sampling bias , I must conclude that Divale ' s statiDtical correlations are spurious . 11 7 Nonetheless , disregarding the methodologica l problems that i nvalidate Diva le ' s work in terms of quantification , and the potential of actual skewing because of indefinite demarcation between age groups , what do the population ratios as Divale has used them show? Table 2 give s selected examples of Divale ' s data , i n the format he uses . This demographic material is i nteresting , when taken case by case . Following Divale ' s reasoning , where males outnumber fema les in the Young Age Group , by inference , female infanticide must be practiced ; when t he males are out- numbered by females i n the Adult Age Group , by infe rence war- fare must be present ; where males continue to outnumber females i n the Adult Age Group, warfare is absent , probably due to European pacification . Looking at the se figure s , we know that the Yanomamo are the type specimen ; ethnographi c data support i nferences made from the demographi c data about t he existence of female infant icide . The Yanomamo also practice male infanticide , but because of the preference to have a male as their first child , they unknowingly kill more fema le s than males . The Yanomamo have only three numbers : one, two , and more- than- two . They are , accordingly , poor stat isticians . They are qui t e unaware of the fact that they do kill more female babies , and every time I questioned them about it , they ins isted that they kill e d both kinds-- "more- than- two tt of both kinds (Chagnon 1968b : 74) . It is not clear at all , however , how many more females than males are ki1led ; Chagnon ' s evi dence is not conclus ive , and he says that hi s demographic data must show cons i derably 1ess fema le ifl~anticide than i s actually practi ced . 11 8 TABLE 2 . DIVALE ' S SE1: RATI O DATA* Males per Sample Year Total 100 Females Number Group of Popu-Census l at ion Young Adult Age Age Group Group 6 Cherokee 1 71 5 94 98 8 Cheyenne and Arapaho 1880 11 4 112 14 Cree (Alberta, Saskat che - 1894 1,000 106 76 wan , Montana ) Cree 1899 1,000 95 96 Cree 1904 1 ,000 100 87 Cree 1919 1 ,000 1 01 88 Cree 1934 1 , 000 93 93 16 Crow Agency 1880 11 5 88 1 7 Eski mo , Alaskan 1839 6 , 560 103 106 18 Eski mo , Caribou 1929 145 80 23 Ki owa , Comanche 1880 106 72 34 Zuni 1935 2,036 123 136 Average , 34 North American Groups 11 8 86 Cashi nahua ( 3 villages ) 206 35 Bal ta 1966 136 70 36 Samue l 1966 237 92 37 Sika 1966 11 0 86 38 Yanomamo 1968 400 129 102 Average , 6 South Ameri can Groups 890 165 88 47 New South Wal es 1846 11 8 1 55 53 Murngin 1930 3 ,000 143 70 55 Tiwi 1928 109 84 84 Average , 1 1 Australian Groups 10, 258 1 31 11 2 59 Kapauku 1955 181 146 81 73 Maring (Tsembaga village ) 1963 204 148 11 5 Average , 17 New Guinean Groups 423 , 348 174 142 11 9 TABLE 2 (Continued ) Males pe r Year Total 100 Females Sample Number Group of Popu-Census l at i on Young Adult Age Age Group Group 77 Gaze lle Peninsul a and 1 9 14 27, 853 124 11 4 adjacent i slands 80 Vitu I s l ands 1 9 14 2 , 523 149 109 81 New Ir2 l and (Namatanai 1 9 14 11 , 739 11 4 100 District) 82 (Kavieng Dis trict) 1 914 14 ,749 126 1 51 87 Lihir Islands 1 9 14 2, 8 18 147 1 4 1 88 Tanga I slands 1 914 1 , 24 1 126 1 21 Average , 1 3 Bi smarck Archipelago Groups 98 , 399 124 124 Fiji Islands 92 Viti- Levu District 1880 7,236 122 98 93 Tribe II AII 1880 1 , 381 138 98 94 Tri be "B'' 1880 984 127 90 95 Wa i ni mala 1880 1 , 71 9 134 99 96 "Mission " Lau 1880 6 , 708 129 125 Average , 5 Fiji Gr oups 18 , 0 28 130 102 102 Tikopia 1929 1 , 278 136 104 'l1ikopia 1952 1 , 753 11 8 ·105 Great Andaman I sland 103 Aka-Cari 190 1 39 300 106 104 Aka- Kora 1 90 1 96 7 3 96 105 Aka-Bo 1 90 1 48 70 93 106 Aka- Jeru 190 1 2 18 185 122 107 Aka-Kede 1 901 59 1 50 80 108 Aka- Kol 1901 11 300 300 109 0ko-Juwoi 1901 48 700 11 0 11 0 A-Kucikwar 1 901 50 66 22 1 1 1 1 Akar- Bale 1 901 1 9 300 50 1 1 2 Aka- Bea 1 90 1 37 75 8 7 Average , 10 Andaman Groups 625 132 1 1 2 * Extra cted from Divale (1970 : 16- 23) . 120 The Kapauku are the second type specimen , and the eth- nography supports the existence of warfare but does not men- tion i nfant icide practices . Ye t the Kapauku sex ratios show arr.ore dramati c shi ft than t nose of t he Yanomamo . The Zuni i n 1935 were certainly not engaged i n warfare , and there a re many more mal es than femal e s i n both age groups . Thus , we are s uppo sed to i nfer the practice of i nfanticide but the absence or discontinuation of warfare . Di scontinuation i s supported by f re quent Zuni warfare r eported fo r the 1880 ' s . The Tiwi r a tios show no change , and we know t hat female i nfanticide was unthinkable for them , given the high va l ue place d on femal es even before they wer e born in marriage , alliance, and l abor . We a l so know t hat the Tiwi fought , and frequent ly , yet male mortality was l ow i n the se encounters . Here , neither femal e i nfanticide nor warfare i s having any demogr aphic effect . Yet we woul d not know that they did have war on the basis of sex rat ios a l one . Looki ng at the ratios for Tikopia, it woul d appear that they practiced fe mal e i nfanti cide and warfare in 1929 , or something was happe ning to those excess mal es i n the Young Age Group , or that in 1952 female i nfant icide was pract iced less often . But we know from Firth ' s ethnographi es tha t the Tikopia di d not engage i n warfare abori ginally . We know that t he Andamane s e frequently fought each other but not anyone e l se . But the r atios a lone do not t ell us mu ch because t hey are extremely varia bl e from group to group , probably because of the smal l 1 21 population s i ze upon whi ch the ratios are based . The addi - t i on or subtraction of just a few members of one sex to these popi.;_l at i ons woul d dramat icall y change the ratios . My point i s , how trustworthy are the i ~ferences made from demographic figures alone ? For some cases the ethnographie s corroborate the inferences , for others they do not and may even contra- dict the inferences . How are we to know which figures reflect life a ccurately and which do not , wi thout checking out every ratio ethnographically? Problems of int erpre tation a nd i ntervening variables , i n addi tion to those concerni ng t he reliability of raw demographic data from pri mitive soci- e t i es , would appear to make the demographic approach, however desirable , difficult i ndeed . There are 2 othe r c omponent □ i n Divale ' s "warfare syn- drome ": polygyny and alliance through marriage . Where prac- t iced , polygyny "worked as a buil t - i n source of conflict as a result of an une qual distribution of women to i nsure a contin- uation of warfare " (Divale 1971 : iv) . But political a lliance s ini t i ated and maintained t hrough the exchange of women "worked as [ a ] regulatory device in limiting the int ensity of primi t ive warfare " (Divale 1971 : iv) . I have found that marriage rules of primit ive societies tend to be explainable in terms of t heir part icular war- f are patterns . For example, Yanornamo and Kapauku marriage rules favo r more concentrated f l ows of women and as a result build closer bonds between allied groups ..•• Conversely , Ibo marri age rul es almost prohi bit the estab- lishment of strong a lliance pat terns by preventing Ego ' s generation f rom duplicating the marri age patterns of t he previous one .... A tentative conclusion i s that a lliance 12 2 and marriage patterns are i nterrelat ed with the i ntens ity of the warfare ; that i s , groups wi th weak alliances have more intensive warfare s i nce no village or group has really strong bonds wi th another while converse ly, groups wi th strong alliances rare l y f i ght amongst themselves . These conclusions are very tentat i ve however and do not contain the degree of support i ve evidence that the infan- ticide feudi n and ol r ny components of t he syndrome have Divale 1970 : 9-1 0 ; i tali cs added ). He is wi se to be cautious . While the di ctum "marry out or die out tt is perhaps general ly true, there are various ways of sat i sfying that ne cessity, with varying degrees of success i n a lliance format i on . At one extreme , Sahlins (1 968 : 60 ) describes the conditions arising from the most " inbred" alli- a n ce form , that of s i ster- exchange . The politica l community as a whole would be splintered into endogamous pairs who exchange women with and t hus are allied only to each other . Or , what i s more likely , the political community would con- s ist of two halves (exogamous mo ieties ), which f orm a se l f - sufficient whole . "Certain Amazoni an tri bes disp l ay just this form of dual organi zat i on , wi th each village a union of .,_ two i nte rmarrying hal ves" (Sahli ns 1968 : 60 ). At the othe r extreme , such re ciprocal · ex change i s prohibited and lin- eages--especiall y segmentary lineages--be come more important than marriage in politi ca l alliances . What side do you take when your r elatives by marriage dispute with your relatives by l ineage? ... The contra- diction is endemi c i n lineage systems , but it can be mi n- i mi zed by a "prohi bition of parallel marriage "; that i s , a bar on repeated unions between the same lineages (Sahlins 1968 : 62 ). 123 Consequently , a l ineage has affinal t i es to several other l ineage s , but while the absence of i ntense re ciprocal exchange may create a wider network of marit a l a lliance s , t he ties are weaker , a nd allegiance to lineage i s preeminent . Nonparallel marri age lends i tse lf to a loca l consensus of belligerence toward eve ry one else . In case of dispute with a no the r lineage , at be st only one of the loca l fami - lies, directly conne cte d by marri age with that group , might have reservations about the propriety of fini shing them off (Sahlins 1968 : 62) . With these extreme s in mi nd , le t me reconsider Divalers assertion that Yanamamo and Kapauku are cases i n which the patterns of woman-exchange create stronge r a lli- . ances than among the I bo . Firs t , note that all 3 engage in fre quent or continual interna l warfare . Second , the Yanomamo case , Divale t s paradi gm case , i s rather complex i n patterns of social and politica l organizat ion . Yanomamo villages are theoretica lly comp osed of 2 unnamed , exogamous patrilineages , and marriage i s based on brother- sister exchange . Me mb e£s of a lineage call all other members of the same sex "brother" or "sister ," and mal es of one lineage call a ll male s of the other lineage "brother-in-law " and a ll fe mal es "wife ." Therefore , a ll "brothers " are in competit ion with each other fo r t he "wives ." If a third lineage should j oin a village , whi chever of t he exi sting lineages in t he orig i nal dual organization e s t ablishes brother- s i ster exchange will find i ts tie s to the othe r or i g i nal lineage corresp ondi ngly weak- ened . Because of the se shifts , and because of t he 124 competi t i on among "brothers " manifested in club fights over possessi on of woffien , villages readily fiss i on . 0fic e this happens , even though the Htales remain agnates , vio l ence reigns . The most bitter fighting , i n fact , takes pl a ce betwe en members of di fferent villages who are rel ated to ea ch other agnat ica lly . I t i s no t uncommon for men to kill their classi f icat ory brothers i n the ensuing r a i ds (Chagnon 1968b : 66 )o Re s i dentia l units cont i nue to fission as generat ions pass and population grows to the maximum tolerable l eve l of about 200 people , or men fight over adul tery , rights to marriage , and murder . Although affinal ties ( reinforced through brother- sister exchange ) seem to be stronger in the long run t han agnatic ties , matters of polit ica l a lliance do not proceed s moothly , either . Yanomamo villages do f orm polit ica l aJ.li- a nces with other villages , "negot i at i ng" t hrough 3 stages : trading , feast i ng , and finall y the exchange of women . The Yanomamo tend not t o attack those village s with whom t hey trade and feast , but they will do so , e specially if "s ome specific incident , such as the abduction of a woman , pro- vokes them" (Chagnon 1968b : 98 ) . I f revenge is taken for abduction , then abduct ion. itself must be carri ed out i n the first place by one member of the embryoni c all i ance . At any point i n t he s tages of a lliance formation , hos tilities may break out . While the thi rd stage , the exchange of women , does link vi llages together through affinal re l at ionships , it 125 appears to me that what may hold t he s e groups together is not completed. exchanges but "the obligation to each othe r to co11- tinue t o exchange wome n " (Chagno n 1968b : 98 ; italics ae;.ded ) . Part ie s to an alliance are very r eluctant to enter i nto the third stage of obligation : weak villages may be afraid that they will not receive a s many women as they gi ve ; strong v illages for ce weaker one s into unequal exchange and do receive more women than they ce de . The weak , therefore , are compelled to exaggerate their strength by bluff and intimida tion and by attempting in general to appear to be stronger , mil itarily , than they really are , thereby hoping to convince their partners that they are equals, capable of i ndependent existen ce . By so doi ng , they also inform their partners that a ny attempt to coerce them out of women will be met with the appropriate reaction , such as a chest - pounding duel or club fight [which can e s ca late i nt o war] (Chagnon 1968b : 98- 99 ). While this may be the ideal pat tern of a lliance format i on , Chagnon notes the followi ng di screpancies . Rarely doe s [the ideal pattern] develop far enough to reach the stage where wome n change hands , part icul arly if t he t wo vill ages concerned are of approximately the same military strength . Fights and argument s over women or food develop , and the principa l s withdraw temporarily on semi-hos tile terms, perhaps attempting a rapprochement sometime in the future . Or , i f the principals are obvi- ous .ly different military potent i a l , the s tronger of the two will coerce it s we aker partner i nt o ceding women early i n the alliance development , taking a dvantage of it s own military strength , t hus altering the course of alliance development in the oppo s ite direction (Chagnon 1968b : 99 ). . Tha t i s , host ilities will begi n again . Naroll , i n his cross - cu l tural quantitative study of the deterrence hypot hes i s , discussed more fully i n Chapter I I , based on hi s small 126 sample fou.n that t he exchange of women di d not r e duce the possibility of hostilities . Unfortunate ly, nowhere does Naroll provide the frequency di stri but i ons a nd statistical corre lations fo r his conclusions i n r egar d to the hypothes i s that cultural exchange manifested in subs idie s , trade , and women reduce the likelihood of war . Wome n--whether they are purchased , wooed , or r aped-- become part of t heir husband ' s househol d ; the exchange of women , therefore , repre sents the most complete form of cul tural conta ct . I n many primitive groups men seek wives in communities which are a l so their potential military foe s . Our compari sons , regret ably , show no significant rela- tionshi ps at all between the freque n cy of war and t hese t hree measures of peaceful i ntercourse (Naroll 1966 : 20) . The exchange of women cert ainly has not l essened the like li- hood of fi ssioning and conflict among Akwe- Shavante groups , either (Maybury- Lewis 1967 ). Nor do marital ties prevent the breaki ng of alliances among t he Du.gum Dani of New Guinea (He i der 197 1 ) . There i s great conflict between moieties i n each Shavante community and fact ionali sm i s not prevented by s odalit i es , a lthough warrior age sets exist and are important i n Shavante life . So, where alliance t heoreti cally should be stronges t due to recipr oca l exchange of women , there are a number of condit ions to prevent or nullify such a harmonious relation- shi p . Why are Yanomamo (and Shavante ) a lliances so fragile ? Why are political r e l ationshi ps withi n and between Yanomamo villages so subject to di sarray? Do we have a case of 127 natural ma n , a Hobbesian creature of every man ' s hand against every ot her man? Wl1at do the Yanomamo value more than alli - ance? Well- documented by Chagnon , it i s belli co s ity : the i mage and awareness they have of themselves . They would rather avoid full commitment to all iance because it goes agai nst their self-image of i ndivi dualist ic se l f - sufficient bellicosity . Inst ead of a ccepting Divale ' s cont ention that war fare i s a consequence of weak alliances, plus other condi- tions , one coul d argue equally we ll that weak alliances are a consequence of warfare--its frequency and t he expe ctat i ons and att i tudes the participants have towards i t . Elsewhere, Divale (1 971) gi ves a general "description" of primi tive warfare , i n which he states t hat " •.. primitive so cietie s throughout t he world lived more or l ess in a state of perpetual warfare , " and" ... s i nce primi tive warfare occurred almos t universally, i t i s an i ndicat ion tha t warfare served an i mportant function i n the cultural - e cologica l adap- tation of pr i mitive cultures , " and it was so important t hat it " ... pervaded a l most every aspect of social and i ndividual behavior . ... Thi s i s to say that warfare was a normal condi - tion in pri mi t ive culture and di d not represent disequilib- rium " (1 971 : iii). That i s , warfare i s a normal condit ion where bands fight other bands and t r i bes fight other t r i bes . But where chiefdoms , say , fight tribes , ... primitive warfare in such an i nstance was the result of t he disequilibrium that resulted from the cla s hi ng of cultures of varying complexities . This type of warfare 128 co ul d not go on i ndefinitely and would end when a new equi libr i um was est abli hed ; when , for exampl e , the retrea t of the tribal s i nto a remote area that i s not e cologically expl oitable by the more dominant chiefdom (1971 : i i i) . At thi s point , I share Aberl e ' s di smay a t t he use of t he equi libri um model : when is equilibri um di sequilibr i um and when i s i t equilibrium? I t woul d depend on whose s i de you were on . Surely f or the "tri bals " t o be saved f rom furt her losses only be cause they retreat i nt o l ands the wi nners have no use for woul d pl a ce t hem on the shor t end of an "equilib- riu.m ." As Aberl e puts the matter , "Concei vably ail di sequil- i brating relationships be t ween t wo s ocial organi zat ions can be shown to be e qui l i brat i ng devices for one of t he t ,:,rn par- ties i n the conflict , but it seems doubtful t hat they can be seen t o be equilibrat i ng for both parties " (1 968 : 99 ). A sim- ilar confusion exists i n modern col d war politics , where one side re cognizes t he exi stence of a "balance" of power only if it has the edge i n mi litary superi ori ty . Wi thout hi s specif ica lly ment i oning i t , it a ppears that Di vale ' s general izat i ons to the uni ve r se of primitive societies in his 1971 essay are again .based on " type speci- mens " : the Kapauku , the Yanomamo , the Central Eski mo , and Plains I ndians . 14 For example , the fo llowi ng generalization is obv i ously based on t he Yanomamo materi a l, whose "standar d pro cedure " for treachery I f i nd t o be unusual, even i n Sou th America . 129 As wit h all warfare , trea chery was an acceptable strategy of p::ciic"i tive societies . A standard procedure was for one group to i nvite anothe r to its village or camp for a feast . I n primi tive polit i cs i nter- group or i nter- village feasting was a proce ss by which alliances were built . Even if t he group invited to t he feast suspected a doubl e-cross , t he y would usually a ccept the invitation be cause to r efuse mi ght i mply fear , which in tribal poli- tics was an open i nvi tation to be attacked . A third group i n treacherous compli c i ty with the host village woul d wait i n hi di ng , to attack when the visitors were drunk or sleeping or about to return home . Sometimes the ho s ts themselves would s uddenly turn on their unsuspect - ing guests . Tre a chery usually resulted i n t r eme ndous slaughter because the vict i ms were unaware and i n clo s e physical proximity to their kill ers . It was not uncommon f or mo st of the vi s i ting men to be murdered and most of the women to be stolen ( Divale 1971 : vi) . I n a footnote (1 971 : vi), Dival e supplie s 3 a ccounts to sup- port this general i zation : The Yanomamo , the Roman rape of the Sabi ne women , and the story in Ge nes i s of Si me on a nd Levi , sons of J a cob , who took revenge on t he non- Hebrew seducer of the ir sister , Dinah (Genesi s 34 : 1- 31 ) . Divale sees the "purpose " of primitive warfare t o be revenge , most ly for adultery and wife- stealing , but a t the same time he de clare s that " ... i t was rar e for a raid to be conducted solely t o steal women " ( 1971 : viii). He further associa te s types of f i ght i ng ( feuding , raiding , pitched bat- tles) wi th popul a tion densities (bands , l ow- density tri bes , and high- density tribes , and some chi efdoms , respe ctively ) . These associat i ons are a l ready fami liar to the reader , but base d on the variable of economics (Newcomb 195 1, 1960 ) . Dival e specifically rejects economic s as s i gni f icant i n primitive warfare. 130 Primitive warfare was conducted for purposes of blood revenge and not for economic mot ives .... disputes over women were the major c ause of primitive wars (1971 : viii) . Many of Diva le ' s (undocumented) generalizations are familiar and testable , but Diva le i s a functionalist and determinist and offers us , in the end , a unitary theory of robot man . Severa l Anthropologists [sic] report that charges of wi tchcraft and sorcery were also the cause of many primi- tive wars . This notion , however , probably confuses cause and effe ct . As many primi tive societie s do not believe that death can occur from natural causes ... , whenever a death of thi s type occurs it is usually charged to wi t ch- craft and sorcery . The kiusmen of the dead person natu- rally ask "who was the wit ch?" and the response , almost i nvari ably , i s an enemy of the dead man . The · relevant point i s that charges of witchcraft and sorcery were usu- ally directe d against individuals or groups where prior disputes were present (1971 : viii). Well enough and good . There i s a vast body of data to sup- port t iiis straightforward conclusion, both on the part of the p9opl e and on the part of Divale . But Dival e then suggests 11 ••• that witchcraft be viewed , i n respect to warfare , as a mechanism for maintaining group solidarity and hate for the enemy, rather than as a cause of primit i ve warfare " (1 971: viii), at which point I throw up my hands iri despair at the futility and dehumanizat i on of functionali st-ecological . 11 explanations . 11 Divale 1 s restatement ·of Ardery/Sumner/ Hobbe s ' amity-enmity complex simply substitutes witchcraft for warfare as the me ans to achieve soci al cohesion , with even less success than Murphy ' s use (1957) of that hypothesis to explain Mundurucu warfare . 1 31 4 . Wa r and the exercise of powe r . War is a n evil , i s a proposition so familiar to everyone that i t woul d be tedious to deve lop . No one i s forced to engage in it by i gnorance or kept out of it by fear , if he fancies there i s anythi ng to be gai ned by i t . I sup- pose that no one will di spute that We went to war at first i n order to serve our several interests ; that we are now , i n view of the same interests , ciebating how we can make peace ; and that i f we separate without having as we t hink our rights , we shall go to war agai n . --Thucydides "Now what I have t hought , " said Arthur , 11 is this . Why can ' t you harne ss Might s o that i t works for Right? I know i t sounds nonsense , but , I mean , you can 1 t just say there i s no such thi ng . The Might is the r e , i n the ba d half of people , and you can 1 t negle ct it . You can ' t cut i t out , but you might be able to direct i t , if you see what I mean , so that it was useful i nstead of bad . 11 --T_ H. White , The Once and Future King Matters of confl i ct and vio l ence have commonl y been treated as belongi ng to the poli t ica l sphere : l aw re i gns within socie ty and war rages between s ocieties . War i s t he absence of law. Yet there is a definitional oddi ty here . Repeatedly , writers refer to the rules of war ; repeatedly , they separate primitive warfare from modern warfare by descr i bi ng the former as game-like or sportive and , as with al l games , subject to rules and , as with all games , not t he real thing . Note the distinction we make i n our l anguage between wa r games , which the Penta gon or the Rand Corporat ion play , and war . The former i s an enactment of war subj ect to certain rules , but i n the long run , j ust a game . I ndeed , if rea l casua lti e s were to result from war games , a l egal 132 i nve st i ga ion woul d ensue to determine negligence , malicious i ntent , or a ccident on the part of participants i n t he "game ," because one of the rules i s that no one on either side i s supposed to get killed for re a l . Like chess . But there are also rules for rea1 war , which is not supposed t o be a game . On the one hand , " ... i n many so ciet i es of the world , the rule s of warfare are so pronounced as to make t he warfare seem a game ," whj_le on the other hand , "Modern t went i eth- century warfare is , if not uni que , at l east rare i n the sense t ha t i t i nvolves t he total society and an absolute minimum of r ules commonly understood by both s i des--the game element is mini mal i n rtotal war '" (Bohannon 1963 : 304 ; ita lics added ). So we have what seems to .be a continuum : t he more rules t i ere are , the more game-like and les s true i s the warfare ; conversely , the fewer the rules , the l ess game- l ike and more true the warfare . But there are still rules . The c ombatants may totally i gnore them, but never publicl y . Fighting by t he r ules , the Geneva Convent ions for example, is part and parce l of maki ng claim to the jus tice of one ' s cause . Breaking or disregarding tho s e rule s (and be ing caught at it) makes one theoretically subject to punishment . The Plains I ndian war leader who los t any men on a raid i s he l d responsible and must atone in mourning and compensation to the families of t he dead warriors . Within the modern military hierarchy , ea ch officer i s held re s pons ible for t he conduct of his c om- mand , and t he government as a whole discharge s it s debt to 133 the wi dows and orphans with a de cent buri a l and compensat ion . Tha thes e exampl es are nots rictly comparabl e may be attributed to di fferences i n the cent ralization of polit ical power and t he assumpt ion of responsi bili ty for t he death--it is now the enemy ' s fault that a so l dier died , not that of h i s commandi ng officer who followed the rules . There are r ules of warfare in both cases . Wha t else shoul d one be aware of i n studyi ng warfare? Bohannon (1963 : 305) a ssert s the followi ng . The ne cessary requi rement in unde rstandi ng warfare is to note that war is a type of relationship lea ding to a gi v en mode of the mul ti centric system of politica l organ- ization-- the mode ba9ed on violence and minimal communi- cat i on . There are many types of war , as there are many types of counteract i ng i nst itutions wi thin the unicentric systems . There are some place s i n which the war i s dec i ded by warriors in contest , and the ma jor parts of the societies are not concerned . There are wars fought by whol e communi t ie s--mobs in violence . There are wars fought by s pe cialist bodies , ca lled armie s . Most North American Indian s ocietie s were typified by the first ki nd of fighting : t he exploits of warriors were cons i d- ered a dominant va l ue in t he culture ; fight i ng u sually i nvolve d only a few people, and consisted in rather loose ly organized rai ds i n which young men are g iven a chance to shine . They then came home , counted c oup , and were r ewarde with the best women of t he tribe .... the i r purpose was never to a i m at a " just peace " under "our control . 11 15 As a consequence , Bohannon continues , .. . it may be unwise to call t hi s sort of situation [ r a i di ng , heaci-hunting] warfare ; _i_t_ i_s_ m_e_r__ _e _l...._y_a_v_i_· _o_le__n_t s ocia l re l at i onshi · with what seems to us and robabl y to them an ant i socia l sort of expres i on 1963 : 305) . To whi ch I c&n only add , yes--war i s at the very least an "ant i soci a l sort of expression ." 134 But , although there are rules of warfare , we cannot call them laws . The extension of the rule of l aw proportion- ate l y di mi nishe s the likelihood of war , as the elaboration of public l aw withi n society replaces pri vat e law-- blood f euds , vigilantism. The solution to the problem of war offered t i me and again by anthropol ogi sts i n the 1940 ' s and 1950 ' s was to e stablish a single polit i cal unit , a single i n- group . I ndeed , the term i n- group would be meani ngle ss be cause there woul d no longer be an out-group . Here I am speaking strictly of political structures and not of humanistic principles . I n a ny case, ext ernal war may s i mpl y be replaced by i nternal or civil war , rebellions , and revolutions . Somewher e between the pole s of law and war exists the perilous l and of diplo- ~acy, subject to neither the explicit rule of law or the i rplicit rules of war . Using Bohannon ' s ( 1963 : 305 ) conflict model, t he diplomat within a critical but fluid period of ti~e af t er a breach of norm must first establish commona l ity wi t h the potent i a l adversary and the n convey t his understand- i ng to his own group . Unfortunately, i f you do not care for wa r , too often the l atter tas k is more di fficult than the former . Diplomacy is therefore a t hankl ess job . Even when the cultures of the t wo [ political] centers are very close together or even all but identica l, diplomacy is a t hank- l e s s task be cause it gets invol ved with the egoce ntric, ethnocentric i n- group desires and cup i dit ie s of each of the center , as a unicentric group . I ndeed , t hey may unders t and ea ch other only too we ll (Bohannon 1963 : 303) . 135 If diplomatic efforts do fail for whichever of a great variet y of reasons and war does take pl a ce , Bohannon s ees the catastrophe as a means of re - establi shi ng re l at ionshi ps on the old basis , l eaving one wit 11 the chicken- or- the- egg dilemma , or on a new basi s , thereby illus trating a di a lectica l process . Bohannon defines war as "a contest having as it s aim a peace i n which the ba lance of power i s shifted . 11 New Guinean and Nor th American r a i ders do not qualify because relations of enmity with their neighbors are permanent and c ont i nued enmity, not peace , i s t he end to be maintained (1963 : 305) . 11 True warf are ," i f we may call it such , i s a somewhat different business . I t has as it s end peaceful sett le- ment with new political conditions , not continuation of fight i ng . And obvious l y , when we find it, we are going to find societies i n which it •is carried out by specia l - it bodies called "armie s " and others in which it is carrie d out by whole bodies of citizens . I n fact , there i s here something of the same difference as that be tween the moot a nd the court . I t i s , i n f a ct, the diffe rence between a l ynching mob and an army . A lynching mob i s a mode of a community ; an army i s a specialis t organizat ion within a sta te ( Bohannon 1963 : 306 ) . Once again we are f a ce d with tryi ng t o figure out i nto which defi nitional pigeonhole to place the group violence of state- le ss s ocieties : if whole bodie s of cit izens in a state l ess society fight t o achi eve peace , they are wagi ng t rue war . As for t he others , they could be either amateurs or profession- a ls, but t hey are interested in war for it s own sake . It is highly like ly t hat my analys i s of Bohannon ' s criteri a is 136 prejudiced by the Orwellian paradox of waging war to achieve peace, a paradox whose expression in our own time has almost achieved perfe ction , perhaps best of all in The Report from Iron Mountain , a hoax taken seriously for too long a time , which drafted secret cont ingency plans fo r the unthinkable-- that "t rue " peace should break out . I can i mpressionistic- ally conce ive of a deve lopmental continuum of human conflict : In the beginning , people may fight for defense , then for the sake of war , then for the sake of peace ( in Bohannon t s terms) , but then again for the sake of war--wars not as a means to peace (under different conditions , of course) but as a means to more wars . If such i s the case , so much for "true war- fare " in modern times : instead of war being an instrument of foreign policy , foreign policy becomes the ins trument of war--or its specialists . I n careful appraisal of this intuitive curve, one must consider first th~ base upon whi ch political power rests within a society . I accept, with some i mportant qualifica - tions, the paradigm of Harri s , Fried, White, and Sahlins , i . e., access to basic economic resources . Using Fried ' s (1 967) type s of egalitarian, ranked , and stratified soci- eties , and speaki ng at least typologically if not evolution- a.rily , band societie s are egalitarian in the sense that everyone has equal right to tools and natural resources . If they are poor , they are equally poor and may not nece ssarily be so (cf . "What Hunters Do For a Living , or How to Make Out 137 On Scar ce Re sources ," by Richard Lee , 1968 ). Or der i s main- tained. i n s uch s ocietie s through t he absence of superi or fire power , t hrough the option of movi ng away from potential c on- flict , through the t hreat of adverse public opinion and thus ost raci sm or death , and t hro~gh t he need for cooperation in all vital human act i vities . Of the se , public opinion- -the requi rement of being a ccepted as a "r egular guy" by one ' s fe llows--is fasc i nat i ngly effe ct ive , even i nto our own t i me , whether t he non-conformer be called "witch" or "communist" or 1tlibber . 11 The a ccused heretic i s everywhere damned , although i n compl ex societies t his trivial truth may be ob s cured by the sheer weight of law and its administrat ion . Everywhere , the prose cut ors of non-conformers--whether shaman , divi ner , priest , policeman , or judge--wi e ld f ormidable political power , albeit somet i mes unoffici ally. As for officia l polit- ical leaders of ega litari an soci e ties , Harris caricatures their powe r : Both headmanship and egalitarian chieftainshi p are likely t o be frustrat i ng and i rksome positions . The cumulative i mpressi on con j ured by egal itarian chiefta i nship a mong Brazilian I ndian group s i s tha t of an overze a lous scout- master on an overni ght cookout . The first one up i n the morni ng , t he headman/chi ef trie s to rouse his c ompanions by standi ng in the middle of t he village plaza and shout- i ng , "Everybody up for the fi sh- poisoni ng expedition ! Let ' s ge t those women i nto the manio c gardens ! Anyone for roof- that chi ng? " The egalitarian chieftain seems to cajo l e , harangue , and plead from morni ng to n i ght . If a task needs to be done , i t i s the hea ' man who starts doi ng it ; and it is the headm3.n who works a t it harder than anyone e l se . Moreover , not only must t he egali tari an chief or headman se t an example for hard work , but noblesse oblige , he mus t a l so set an exampl e for gener - osity (Harris 1971 : 385 ) . 138 This headman has no means to phys i ca lly force anyone to do anythi ng, although he mi ght use sanctions of public opinion and threats of wi tchcraft . I f he does not behave as a good headman shoul d , people will s i mply i gnore him and there is not a thing he can do about it . Political power is diffused throughout the social group . For such societ ies , Fried ( 1967 : 101} contends t hat even i f ethnographers have described them as warlike , the time i nvolved i n preparations , batt l e , and ceremoni a l s seems little . Bands do no t build fortif i cat i ons , do not stockpile food and materiel , do not proviQe any special military trai n- . i ng for warriors , and do not possess tools spe cif ically used i n killing men as distingui shed from those used i n hunting . The typical action i s a rai d involving few attackers ; t he appropriate word for what takes p l a ce seems t o be cla sh-- there i s a sudden violent set- to and most of t he part ici- pants return hoarse from screaming threats and insul ts but are otherwise uns cathed . Aga in , i t is tempting t o compare this behavior with that of the other primates [those same primates whom , the reader will re call , Ardrey cons i ders to be "evol ut i onary failures " ] who , we a r e told , expend most of their e ne rgies i n combat s i tuat ions carrying out threatening behavior rather than a ctual ons l aughts . This is not to say that warfare on the egal- i tar i an- band leve l i s devoi_d of casual tie s (Fried 1967 : 1 02) . Fried suggest s that the Siriano are possibl y clos e to the norm of warlikeness of simple egali tari an societies : t hey do not fight with each other and they retreat f r om rather than fight outsi de groups . Other classic s of peaceful bands i nclude Shos honean groups , the Tiwi , the Copper Eskimo , the Mbuti Pygmie s , and the Yahgan . Yet the case i s not clear . 139 Admi~tedly , t he problem is a tough one . There are i ndi - cations , for example , that the Viki ngs feared people whom they knew as "Skraelings ," who seem to have been Eski mos , for these peopl e paddled right out to the Viking ships and attacked the Norsemen fiercely . This also reminds us of the mres ome reputation of some of the Andamanese who are reported to have destroyed many crews of wre cked vessels over the centuries. This probl em is worthy of much more seriouc attent ion than it has been gi ven be cause the nat ure of warfare throws much light upon many problems of comparative so ci - ology . The fina l r evel a tions may be surprising . Some s i mple so cieties of the ethnographic present may be dis- covered through comb i ned archae ological and ethnohi stori- ca l techniques to have previously been more complex , wi th the more complex structures having been among the casual- ties of massive contact (Fried 1967 : 104). Note that the assumption remains that simpl e egali- tarian so ciet i es are peaceful because they have neither the te chnological means , nor organizat i onal principles of com- mand , nor the tacti cs , nor the casualties to qualify as having war . Fr i ed (1 967 : 105) also comment s on the absence of war l eaders , i. e ., "leadership of military affairs ," as extreme l y significant: " .. . every man stands and fights or runs away by himself ." The context of hi s comment i s refu- t ation of the Spencerian notion that centralizat ion of polit- ical power i s derived from war chiefs who will not give up t heir authority once the war is over , an i mportant point to which I shall return shortly . The probl em of the moment i s what is meant by "war l eader"--wh t does he do , when does he do i t , and do the followers obey? I n almost every society surveyed for the cross- cultural research I have done , i f a people engage in war , there is someone des i gnated as 11 1-mr I 140 leader ." An exception are the Trumai , a depopulated society i n a state of collapse , whose 43 members were no longer ab l e to organize f or anything, even subsistence a ctivit i es (Murphy and Quai n 1955 ) . Furthe :cmore , a war l eader may l ead in at least one and possibl y two spheres : (1 ) the planning , prepa- ration , and initiation of fighti ng ; ( 2) giving command duri ng the a ctual fighting . The crucial di stinction i s to what ext ent i s his l eadership one of authority, i . e ., backed by force . Do warriors obey or do t he y not? Although our que s - tions were phrased somewhat differently , Otterbe i n (1 970) and I have searched the data for answers . Somet i mes we are not i n agreement . For example , Otterbein codes the Papago a s having a high degree of military subordination , i . e ., 1•~.-mrriors obey orders gi ven by l eaders ," as opposed to 11 ~·rarriors frequent ly do not obey orders given by l eaders " (1970 : 144), while I have coded the Papago as havi ng a low degree of subordinat ion , i . e ., "an i nformal l eader whom people obeyed because of respe ct but who had no means to for ce warriors to obey ." Possibly Otterbein and I are aski ng different questions of the data , in which c ase the comparison i s irrelevant . The variable I am trying to measure , as di s- cussed further on in Chapter II, i s the extent to which men voluntarily participate i n fighting a ctivity . One point that does seem to be clear i s that the se simple egalitarian s ocieties , when described by ethnogra- phers , existed at the ends of the earth or i n refuge areas . 14 1 Most of the egalitarian societies known to ethnography have been located in areas remote from the centers of complex cultural development over the past 5,000 or 6 , 000 years or more . Particularly as we approach our own time , we find egalitarian societ ie s in relatively i mpoverished habitats , frequently extreme in climatic conditions and poor in natural foo d resources (Fried 1967 : 110 ). Do they exist in such regions because they have been pushed into them by more aggressive societies? What is involved in the "pushing" or "being pushed"? There are ... a few societies where men seem to find no pleasure in dominating over , hurting, or killing the mem- bers of other societies , where all they ask is to be at peace and to be l eft in peace . These societies are , of course , small , weak , technologica lly backward , and J_i ving i n inaccessible country ; only so could they survive the power- seeking of their uninhibited neighbors •.. • As far as the history of these small tribes can be reconstructed, they have always chosen to retreat into ever more inac- cessible c ountry rather than stand their ground and fight with invaders . There is no reason to suppose that their psychological or physiological potentialities are differ- ent from those of thej_r mor e aggressive neighbors , but their values certa inly are ; for them peace and the absence of quarreling and jealousy are far more important than a reputation for bravery and virility ( Gorer 1968 : 34- 35 ; italics added) . _ Out of such reputations are prestige statuses made , which one encounters in rank societies, Fried ' s second polit i cal type. These societies are characterized by equal access to tools and natural resources necessary to sustain life , but also by differential access to posit ions of prestige , posi- tions '' ... somehow limi ted so that not all those of suffi- cient talent to occv.py such statuses actually achieve them " (Frie d 1967 : 109) . Unlike the simple egalitarian societies , 11 ••• rank societies have managed to come down i nto our own historica l period still i n occupation , if not control , of 142 some of the most desirable stretches of the earth 1 s surface " (Fried 196 '7 :11 0) . Rank societ i es are usually food- producers (with the exception of the Northwes t Coast I ndians ), have l arger and denser populations collect ed i n permanent , aut ono- mous , and significantly exogamous villages organized ac cord- ing to formal kinship t i es of descent--not ne cessarily uni- lineal--where r edi stribution of go ods is more i mportant t han reciprocity in economic i ntegration and the ro l e of village re distributor carrie s prestige and politica l status but not politi cal authori t y (Fried 1967 : 110-19). Kinship i s of pri- mary i ~portance because it i s the organi zi ng principle of a core group whose re sidence patterns are male-domi nated (viri- avru.ncu- patri loca l) l ineages and clans . All the sources of i nterpersonal conflict f ound i n egali- tari an society persist i n rank society , as i nde ed they persist in all subsequently evolved types of society . Ce rtain kinds of irr i tation no t present i n egalitarian society make their appearance i n rank societie s , although their expression may still be relat ively subdued. For example , while a ccess to basic resources within t he cor- porate unit is not s i gnificantly a ltered , t here t eLds to be much more consumer ' s property i n rank s ociety . Pat- terns of reciprocal exchange do operate to keep these things i n circulation, but there is a qualitative break wi th egali tarian soci eties as accumul ation of nonstrate- gic values is often the basis or means of va lidation of rank dist i nctions (Fried 1967 : 141~42) . ·That :i_ s , conflict can arise over who has right to a limited re s ource of non- vital but prestige value . Who has the right to care for and wear certain masks and religious parapher- nalia ; bow much is one entitled to i n an exchange wj_th re spe ct to one 1 s status , in gift - gi ving , in damages ? 143 Obviously, when people are concerne with rank and status , they are at l east as concerned with lack of or t hreat to that rank as they are with re cognition : receiving one ' s just due passes without notice , but anythi ng less breeds t rouble . A line of demarcation between egali t arian and ranked soci eties i s di fficult to draw , as i n all classificat i ons , a nd Frie d readil y a dmits this . He sees ranking a s analogous to biological preadaptat i on : II aft e r stratificat ion sets i n , r anking emerges i n the form of a socio- e conomic cla s s system . As such it s significance i s i mpossible to overest i - mate " ( 1967 : 154) . Actually , ranki ng is signifi cant enough i n differential ac ce ss to pre stige r esources . Fr i ed does not have to explain backwards from strat i ficati on to ranking . As for warf are generally , Fried stat es t hat ... r ank societies tend to be combat ive , ... many of them exi st i n what may be seen as a chronic s tate of war , and . . . terror and psy chologica l warfare are cormnon means of mai nta i ning group i nt egrity i n t he face of compet ition for survival (1 967 : 178) . There are some problems with Fried 1 s i dentif ication of his third t ype , s tra tif i ed s ocieties , because as he admi ts , it is almost impossible to fi nd stratifi ed soci et i es that are not states. Nevertheless , he assumes t hat most i f not all states must have (rapidl y ) passed t hrough a stage of strati- fication in their evolut i on toi.-ards the state . Fried defi nes a stratified society as " ... one i n which membe r s of the same sex and equivalent age status do not have e qual a ccess to the ba~ic re sources that sustain life " (1 967 : 186 ) . The specifi cs 144 of those basi c resources are var iable, of course , according to g eography , technology , and subsistence i deo logy--" the hi s - torically determined perception of t he exploi table environ- ment ," i n Fried ' s words . I n simpler language a minority con- trol a ccess to basic sources of foo d and the ma jority mus t sell their l abor to the minority in order to live . This l abor can be use d i n several ways . The most obvious i s in f urther food production ; the most interesting for our con- cerns is as inst ruments of for ce--police and so ldiers . Why s hould people "gi ve up " political cont r ol over t he ir own lives in the transitions from egalitarian to state soc ie t ie s ? The quest ion has long been a favorite one : Why have pe ople permitted themselve s to be seduced , bilked , mur- phied , or otherwi se conned i nto r elinqui shi ng a condition of egalitarianism for one of inequality? The question , of course , i s loaded , not only politically but cultur- ally . Apart from being biased , however , the que s tion is wrong in it s i mplicat i on t hat indi v i dual s f a ce d a conscious cho ice and se lected the alternative of rank society . I be lieve that the evidence i s quite to the contrary and tha t events conc eived retrospectively as ca taclysmic a ctually passed without noti ce until they were ful ly a ccomplished . Rank s ociety grew out of egal i- t a rian soci ety without t he conscious awareness of the members of the society i n which it occurred ; I believe tha t s t ratified society and t he state emerge d i n the same quiet way a nd were i nst i tutionally fully present before anyone fumbled for a word by whi ch to de s ignate them (Fried 196 7:1 82- 83) . Thus , a ccording to Frie d , trans i tion from one type to another " just happens " a t the very leas t , or more f a irly due to spe cific "i nit i at i ng conditions " in the environment : 145 ... population pressure ; shifts_ in customary postmarital residence patterns ; contraction or sharp natural alter- ation of basic resources ; shifts in subsistence patterns arising from such factors as technological change or the i mpingement of the market system ; deve lopment of mana- gerial roles as an aspect of maturat ion of social and ceremonial system . I explicitly reject warfare and slavery as i nitiating condit ions (Fried 1967 : 196 ; italics added ). But how did the initiating condi tions that exert what we must assume to be selective pressures in Fried ' s evolutionary paradigm t ~emse lves come about? While I do not wish to recount Fried' s explanations for each one of these conditions , I do want to consider popu- l ation pressure for 2 important reasons : it has been a basic premiss in the previous discussion of Harris ' and Divale ' s hypotheses , and it i s a variable i ntercorrelated in my study . Fried is faced with the difficult task of reconstruct i ng events i n the original stratified societies in the Old and New World centers of subseg_uent urbanizat ion, on evidence that is virtuall y nonexistent . Acknowledging this problem, he then analyzes stratification where it has occurred as a consequence of contact wit h more complex (and usually Euro- American) societies . His exampl e is Tikopia , under contact conditions in which indigenous population controls--late marriage , l imite d births , and warfare--were eliminated by Christian missions s o that populat ion between 1929 and 1952 increased by 35 per cent and into famine (Fried 1967 : 199 ; Firth 1959 : 53) . As a consequence , access to land , the basi c re 0ource , became circums cr ibed by more rules based upon 146 membershi i n smal l agnatic lineages , and something like rent had to be paid by non- members for l and usage (Fr ied 196 7 : 200- 201) . Formul ated as a generali zat ion , then, l and s car- city develops due to population pressure , land tenure rule s be come more rigid , although the rules usually delineate right to use r a the r than right to own . Fried does not explain what happens to t he have- nots , or how delineation of land- holding rules becomes hierarchical, which , after all , i s what strati- fi cati on i s a ll about . It might be i nt eresting to reorder Fried ' s i nitiating conditions i n a ccordance with Dri ver ' s and Massey rs ( 195 7) ~ell- tested evo lutionary order of change : division of labor, residence , l and tenure, de s cent , and cousin terms , but con- trolling for geographical region and language family . As Driver (1966), Jorgensen (1 966) , Chaney and Revilla (1 969 ) have shown , in the long run geography and his tory are more significant than uni versal psyches and functions . Tenta- tively , I offe r a revised scheme that integra te s Friedts i nitiating conditions of stratification with Driver ' s and Massey 1 s sequence of change a nd supplie s s ome missing links : [basic resources ]; "contraction or sharp na tural a lterat ion of basic resources "; [technology] ; ( di vision of l abor ) ; "shifts i n subsi stence patterns ari s i ng fro m such factors as techno l ogical change or the i mpingement of the market sys- tem"; ( residence) ; " shift s in customary postmarita l residence patterns "; "popul at ion pressure 11 16 ; (land t enure ); ( descent ); 147 ( cousin terms ) ; [kinship- based stratification]; "development of managerial ro l es as an aspect of maturation of social and ceremoni al system"; [property- based replacing ki n- based st ratification] . Following Fried t s reasoning , the control- ling minority of a strat i fied society woul d subsequent ly use force as a means to solidify their gains , maintain i nternal stratification , and enhance their property , pos ition , and power . I t i s a truism that many , indeed most , people do not rea lize a change has taken place until it is upon them , and that possibly under human conditions of short life expectancy , non- literacy, the habi t of custom ("we have alvrays done it that way"), conservat i sm , and even fatalism , the question of choice is alien . Neverthe l ess , some people do choose . Marvin Harris , despite all his emphas i s on eti cs and the l aws of history , offers a hypothetical re construction of events that makes the reasons , i ntentions , and purposes of the to- be-ruling minority very c lear . Thi s is what might happen . I n redistributive economic systems , the redistri bu- tor ( headman , "big man") maintains his prestige-- remember he has no power--by demonstrating his generosity repeatedly i n giving fe as ts , the produce for which he and his kinsmen have worked long and hard. . Each feast must be more lavish than the l as t if the headman i s to se cure his position , which he obviously must want or why go to a ll that trouble? I n the redistri butive pat ern , guest a l so bring food to a feast and 148 the banquet pot , sweetened by the headman, is divide d up equally among the guests wi h spe cia l portions going to kins- men who worked to provide for the feast in the first place . A l arger and l arger port i on of the kitty i s held back for the headman and his ki nsmen . Feasts recur and guests bring their gifts, their attendance perhaps encouraged through the use of public opinion or the threat of witchcraft . More a nd more i s held back until the host ' s share becomes a royal treasury used to purchase force whereby gifts to the headman become taxes to the king , non-payment of which is a crime against the state , a state headed by a royal lineage and administered through a bureaucracy of spe cia lists (Harris 1971 : 392-403 ). There are numerous variations on this theme , but it is incon- ce ivable that at least some of the people do not know exact l y i,,;-hat they are doing and continually make specifi c decisions to a chieve certain ends . As the discrepancy between the haves and the have-nots widens , proportionately more revenues must be used to maintain l aw and order . Even when the state objectively provides the mass of cit- izens with a measure of securi ty and well- being superior to that of egalitarian peoples, the expropr iat i on of the peasant ' s output , the sealing off .of habitats, and the demand for obedience to authority places the governing class in an essentially unstable and vulnerable position. The evolut ionary viability of the state rests i n l arge measure on the perfection of i nst itutiona l structures that protect the ruling class from confrontation with coa lit ions of a lienated commoners . These structures fall into two bas ic categories : ( 1 ) i nstitutions that control the content of ideology ; and ( 2 ) i nst itutions that physi- ca lly suppress the subversive , rebellious , and revolu- tionary actions of alienated individuals and groups ( Harris 1971 : 406). . 149 Harri s (1 97 1 : 406 ff . ) goes on to lis t the i de ologues of pre - i ndust rial state s such as the I nca, Aztec, ancient Egypt i an , Roman , Med i eval European , Or i ental despot i sms , and modern tyrannie s , a background agains t which the United Sta te s of 1974 i s an anomaly, a n aberrat ion , a r elative ly ope n society wi thin which one can see more easily the maintenance of political power of the state i n operat i on . He does not men- tion 2 s t ates for which we have evidence tha t the r uler s knew what they were doing to achi eve the i r goal of centra li zed power and empire , '.I.'lacae l e l of the Aztecs and Shaka of the Zulus . Tlacaelel , half- brother of Moc tezuma I and e quivalent of pri me mini ster , elevated the god of war Huitzilopochtli to the posi tion of most impo r t ant de ity in t he Aztec panthe on . He de creed that a great t emple be built to this god , and that the surrounding city-states be come marke ts from which the Aztec could obtai n food for their god , humans f or sacrifice take n i n war (Leon- Por tilla 1963 ) . One man i so l ated a minor re l igious cult and generated a myst ica l militari s t pat t ern t hat grew steadily i n scal e and i nt ens i ty until the Spani s h Conques t i n 1517 . Three hundred years l at e r , on t he other s i de of the world , Shaka assumed a pattern of tyrannica l government and military expansi on begun by an uncl e and out of it generated an empire . Although i t is not clear that he was not i nfluenced by European contact, he invented military tact ics that quickly decimate d or incorporat e d non- Zul u 1 50 populat i ons and were successful even against the British for a time . His control was absolute and maintained by terror : any transgression, a ctual or potential , was a capital cr i me . Yet his subje cts did not rebel, probabl y because of fear , but a l so be cause of materi al gains and the grat i fi cat ions of su c- cess i n an expanding empi re (Wa lte r 1969 ) . These 2 men were i nnovators , people who designed and executed pl ans to a chieve the ends for which they are remembere d . The Aztec peasants and perhaps even the warrio rs may not have realized what · t hos e goa l s were . The Zulu common persons may not have under- stood the source and s t ructure of t heir rule r t s powe r . Per- haps those who di d , died . I n any case , it makes events more i n e lligibl e to believe that at least some of the people activat i ng institutions knew what t hey were doing as t hey olanned fo r and i nitiated both change and maintenance of the status quo . The determinist alternat i ve i s to believe that as History roll s on , human beings unknowingly f all into wai t ing roles and speak but do not understand lines t hat It has written. There remains more theoretical variation to consider here . Carne iro (1972) offers a determi nistic theory for the or i gin of the state i n which warfare i s neither the e ffect of stratifica tion (as Fried would have i t ) , nor the cause of strat ifica tion (as Andreski would have it) , but a necessary a l though not sufficient condition . Carne i ro ' s rather elegant t heory i s based on the i dea of ci r cums cript ion, whether 1 51 environmental or soci.al, which stiinulates warfare t hat i n turn stimulates a progression toward political coa lescence . Thus , t here is little question that , in one way or another , war played a decisive role in the r i se of the state . Historical or archeological evidence of war is found in the early stages of state formation i n Mesopotamia , Egypt , India , China , Japan, Greece , Rome , northern Eur ope , central Africa , Polynesi a , Middle America , Peru , and Colombia , to name only the most prominent examples (197 2 : 426) . Carneiro sees warfare as a necessary but not sufficient con- dition . The sufficient condition i s cir cumscript ion . For the environmental type , l and used for food-production is bounded by ge ographical barriers : mountai n ranges , dese r ts , or oceans . His examples for comparison are the coastal val- leys of Peru and the Amazon basin . I n the l att er , as horti- cultural populat ions grew , there were vast forested a reas into which l and- hungry people coul d spill , rather t han com- pete with their neighbors over a limited l and resource . Warfare was certai nly freque nt in Amazoni a , but it was waged for reasons of revenge , the taking of women , the gaining of personal prestige , and mot i ves of a similar sort . There be i ng no shortage of land , there was , by and l arge , no warfare over l and (Carne i ro 1972 : 427 ) . The defeated, rather than becoming subject or tributary to t he vi ct ors , could. s i mply move away and open new gardens in uninhabited areas of the forest . Thus , Carneiro concludes , warfare st i mul ated the spread of horticultural tribes throughout· the Basin, who continued to live i n autonomous di spersed villages . Neverthe l ess , he infers that there may 1 52 have been situa tions of soci al circumscription operating in Ama zonia , basing his argument on the now familiar Yanomamo case . While the Yanomamo live in uncircumscri bed rain for- es t , population density a nd consequent strife a re most intens e towards the center of Yanomamo territory and diminish as villages are located closer to the periphery . War i s most int ense in the cent er because those who are land- hungry can satisfy thB demands of population pressure only by taking l and away from other Yanomamo , and those who are in danger of l osing have no place to escape to . As a consequence , we have a situation in which Yanomamo fight other Yanomamo , the intensity of such warfare decreasing as one moves outward from the center , with those Yanomamo on the periphery not fighting any non-Yanomamo . Contrast the expansive effects of s uch pressure wi th the Tiv case , dis cussed at length earlier, where those on the periphery are forced to fight non-Tiv to r eplace lands lost to other Tiv . While subsistence base is somewhat comparable between the 2 cases , the concerns of lineage organization are different . Yanomamo lineages focus on the exchange of women in marriage ; Tiv lineages focus on i nternal soli darity and land tenure . An explanation based upon environmental conscription i s more stra ightforward . I n Ca rneiro ' s Peruvian example , he presents an infere ntial recons truction of political develop- ment fr.om autonomous agri cultura l villages to imperial status as follows . Agricultural land was l i mi ted to 78 small 153 valleys , '' ... backed by the mountains , fro nt ed by t he oea , and flan:i-cea. on either side by dese rt as dry as any in the worl d '' (1972 :427 ). As autonomous villages grew i n size , they fi ssioned , and as availabl e virgin farm l and de creased , int ensity of cult ivat ion increased and marginal l ands were made productive through terra cing and i rriga tion . But at some po i nt [ doubtle ss below Harris ' ceiling on the t e chno- environmenta l carryi ng capacity of the Peruvian valleys] physical s truggle specifica lly over l and within a valley began . With increas i ng pressure of human popul at ion on t he l and ... t he ma jor i ncent i ve for war changed from a desire for revenge to a need to acquire l and . And , a s the causes of ,var became predominantly economic , the frequency, inten- s i ty , and i mportance of war increased (Carneiro 1972 : 428 ) . The out come s of t hese wars were very di fferent from those waged i n Amazonia. Lo sers had no place to escape i nt o, and if they were not annihilate d , they became politically subor- dinate tributarie s to the victors , resulting in supra- village poli t ica l unit s , i . e . , chi efdoms . As popul at ion continued to r i se , putt i ng even greater pres sure on a rable land , competi- tion took place between chiefd oms , with politica l subordi na- tion once again t he outcome of war , and the l arger , more com- plex vi ctor was t hen a ki ngdom, and a valley embraced a state . The l ast and final step was conquest of valley- ki ngdom by valley-kingdom , the wi nner of this Peruvian tour- nament to be the i nternational grandmaste r . The contest was 154 won by the I ncas , who controlled a s i ngle empi re . The aggre- gat i on of larger poli tical uni ts was a ccompani ed by increas- i ng internal compl exity , speci fically the deve lopment of a n eli te to admi nister conquered pol i t ica l uni ts . And it was the i ndi vi dual s who had di st i ngui shed t hem- se l ves in wa r who were gene r a lly a ppoi nt ed to politica l off i ce and assi gned t he task of carryi ng out t hi s a dmin- i strat ion (Carne i r o 1972 : 429 ) . Admini s t rative duties were of the s ort t ha t , i n my e stima- t i on , i ncr ease in effi ciency i n proport i on t o t he t hreat or exercis e of force : mai nt a i ni ng l aw and orde r , col lect i ng t axes , mobili zi ng l abor f or public works . So we have , i n Carne i ro 1 s s cheme , a n upper cl ass c omposed of a mili tary e lite , t he rul e r , a nd his kinsmen supported t hrough t axa t i on of c onquered food- producers ; a l ower class of pri soners of war be comi ng servant s and s l aves ; and a "middle-cla ss " of t hose di spossessed f r om thei r l and by wa r but not themse lves enslaved who moved i nto l arger communi t i es and sold their labor as workers or a r t i sans to t he upper c l ass and re ce i ved tax g oods as payment . Areas of circums cribed agricultural l and e l sewhe r e in t he worl d , such as the Va l ley of Mexico , Mesopotami a , the Nile Valley , and t he I ndus Valley ,· saw t he pr ocess occur i n much the same way and f or essent i ally the same reasons . I n t hese areas , t oo , autonomous neolithi c vil- l ages were succeeded by chi ef doms , chi e f doms by kingdoms , and kingdoms by empires . Tshe l ast stage of thi s develop- ment was , of course , the mos t i mpressive .... But , in a sense , empi res were mere l y t he l ogica l culmination of t he pro cess . The really fundamenta l s te p , the one t hat ha d t r i ggered the ent i re train of events that l ed to empi r e s , was the change from vi l l age aut onomy to supravillage i ntegrat i on . Thi s step was a c hange i n ki nd : everything t hat fo llowed was , i n a way , only a change i n degree (Carneiro 1972 : 429 ; i talics added ) . 1 55 And apparently this quantum political leap did not take place until about 5000 B. C. Under pressure probably of the myst i que of unitary theory, i n order to account for exceptions Carneiro extends the range of his theory to i nclude chiefdoms that existed a l ong the Amazon , arguing that they deve loped because of resource circumscription . Carne i ro uses the term 1t resource concentration" and says that it 11 arnounted almost to a kind of circums cription" (1972 : 430), whether the resource be the basis of a food-collecting (such as the Northwest Coast tribes) or f oo d-produci ng economy . I n summary, Carneiro t s theory simply states that political centralization comes about through warfare relative to the scarcity of resources due to geographic or social conditions of circumscription. I f Otterbein 1 s 1970 cross-cultural results are reli- able, they seem to l end support more to Andreski and Carne iro than to Fried : a centralized polity c annot persist without the backing of a sophi st i cated military force, but a soc i ety can be efficient militarily without political cen- tralization (1 970 : 70-76) . I n his Fore word to Otterbein (1 970) , Carne iro i s most i mpressed by Otterbein t s observation that i n war the test of fitness is applied , not just to military practices but to so cieties themse l ves . And , Carne iro continues : The ultimate test of fitness , of course, i s survival. And no matter how we l l adapted a society may be in other 1 56 respects, if it proves unable to cope with its enemies it has failed i n i ts over- all adaptation and must give way . When societ ie s fight, the cultural equival ent of natural selection comes into play .... As societies compete , the less well adapted tend to fali by the wayside , leavi ng outstanding those best abl e to witts~and the compet ition. From the point of view of the trait s involved , ... more efficient traits survive and spread , while l ess efficient ones decline and disappear .... Cultural se l ec t ion , which operates even on trai ts of l ittle or no adaptive value , acts with special intensity on traits directly concerned with survival . And since there i s generally no greater chall enge t o a society ' s existence than war , i t is here tha t we find se l ection operat i ng most rigorously ...• Coldly viewed , warfare has enormous ecological signifi- cance . It i s concerned , after all , wi th a most vital aspect of a society ' s environment-- i ts enemies . I f waged successful ly , war means the preservation of a society ' s i nt egrity and independence , and the defense or even increase of it s territory and resources. If waged unsuc- cessfully , warfare may mean defeat , subjection , or even exte rmi nation . This i s ecology in spades ! (Otterbe i n 1 970) . On ce again , and even more s i gnificant ly since what i s subject to evolutionary processes here i s the very polit i cal survival of a society, we must conclude , including Otterbein ' s 1970 findings, about which more l ater , the following! Those organisms best adapted to a g iven environment will tend to live longer and produce more off spring than those organisms l ess well adapted . Those cultures with the most efficient energy-capturing sys tems will tend to inc~ease and spread at the expense of those culture s with less effi- cient energy-capturing systems. Those political communities that can capture, maim , and kill their enemies most 1 57 efficiently will tend to expand i nt o the te rritories of thos e political communities with less military efficiency . And we know 1:rhich organisms , cul tures , and political comm.uni t ies are be st a dapted , most efficient , and most successful be ca use t hey reproduce more , spread farther , and kill better . Like one lost in the woods , the eco- functionali sts have l ed us around in a cir cle . Answers "Have you guessed the r i ddle yet ? 11 the Hatter said , turni ng to Alice agai n . "No , I g i ve it up ," Alice replied . "What 1 s the answer? " "I haven ' t the s lightest i dea ," sai d . the Hatter . "Nor I," sai d the March Hare . Alice sighed wearily . "I think you mi ght do something better wi th the time ," she said , "than wasting it in aski ng riddles that have no answers ." - - Alice i n Wonderl and Like Alice, I am weary of looking f or an answer t o a riddle that may not have one . At least , there may be no s i ngle answer , no one test ab l e gene ralization that will cover all cas e s . I n order to be cons istent with the pri nciples of theoret ica l validity set forth earli er , it is highly unlike l y that t here i s any singl e set of generali zations that can be subsumed uncler a law-like statement when we are deali ng with human behavior . At the same t i me , i f I fo llow my c onception of human action to its l ogical conclusion , t here is the danger of e ndi ng with a description or map on the s ca le of 1 1 58 to 1 , a condition of coming around full circle and dupli- cating t he world , at the very least a physical impos s ibil ity . So some~here between hither and yon , now and t hen , micro and macro , l i ght and dark , there i s a l and of meaningful unde r - standing . Perhaps Harris and others like him woul d ident i fy such an area as that of "middl e-range " t heories and mushy ones at tha t , but I contend that this i nterstice hol ds explanatory power that may be closer to real human be i ngs acting i n a real worl d . This interstice has been vari ously l abe led , and often dismis sed , as " semantics " and "cognit i on ." Chaney (n . d .: 18 ) , i n his concern to return dest i ny t o human control , makes a similar cl a im for meani ng : Both t he unique and universal i n the space-t ime- mode- meaning-significance continuum of human exi stence are to be expla ine d i n terms of the emergent quality of man 1 s pan- human me aning- medi ated existence . What we ca l l a cult ure , society , or "periodization " i n human h i story i s better conceptualized as an artificial, hi storica l ( spatiotemporal) clustering of l ocal distortions in the space- time - mo de - meaning- signifi cance cont inumn of human existen ce . This view appears to be consistent with that of Bi dney , that because culture is an open system subject to human se l f - awarenes s , it i s i ndetermi nate ( 1967 :1 7-1 8 ). I must take i ssue , however , with Chaney ' s descript ion of a l l cultures as artifici al and distorted , because one does not know what would subsequent ly be real and undistorted . Patterns do exist in human act i ons . As indivi dua ls we t end to behave with some measure of cons i stency . We sha re with our neighbors and our children s i milar but not identica l 1 59 ways of a ccomplishing daily tasks i n everything from whe n to i nt roduce soli d foods to an infant to how to govern our com- muni t ies . Our lives are gui de d by norms and standards agai nst which we continua lly measure ourse l ves , neighbors , children ..• and stra.ngers . Chaney (n . d .: 18 ) conceive s of sociocul- tural patterns as ... hol ding in terms of the continuity of~ culture (habitua tion and conditioning of human , art ificial para- di gms for expectation and a ction differentially re i n- for ced t hrough t he satisfaction of primary and derived "urges "), rather than in t erms of s omething inhe r ent in sociocultural phenomena analogous to "c osmic glue ." Thus , we learn to behave i n certain acceptable ways and are taught by other human be i ngs . I have some doubt that teach- i ng and l earning are suitably explained in behaviori st ic ter- mi nology , but that i s not a n i s sue here . Again , Chaney 1 s view appears to be consistent with Bidney ' s (1 967 :1 7-1 8 ), 11 that some degree of cultural dete r minism characteri zes human life , a nd t hat cultural anthropol ogi s t s may study compara- tively the s ignificant corre lat ions between culture, person- ality , and s ociety ." If t he reader will recall from my ini- tial discuss ion of determinism , i t is not at a ll clear that culture i s indeterminate but that a human withing cul ture i s det ermi ned . Yet Chaney ' s di stinct ion mus t be observed : it is only what people learn from ot her peopl e in a particular group that can be spoken of in some measure of determinism , _go t that cla sses of sociocultural phenomena i nherent ly and therefore uni versally hang toget~er , like magnetism or 160 gravi ty . I mus t ob ject, howe ve r , to the notion of "art ifi- _Qial paradigms " and instead offer Spra ley ' s and McCurdy ' s (1 97 1 : 4) conception of culture or its paradi gms as "arbi - t rary ," i . e . , there are many cultural solut ions to a problem of human life . 17 This term avoids the logica l pitfalls of both "artificiality" and re l a tivi sm . I speak of i nterst i ces , Chaney (n . d .: 19 ) speaks of transcendance : I wish to stress tha t the present di s cuss ion attempt s to trans cend disjunctions such as absolute- relative, i dealis tic-materialistic , rational-empirical , free will- determinism which erode i nto ( 1) searches for .the direc- tion of t he causal arrows or (2) anything goes . Rather , the st ress here i s that human beings exi st in var iously interre l ated conceptua l plots which mediate their exist- ence . So in the end we again face the old ultimatum of e ither order or chaos, find both l a cki ng , and seek a third alternative . At this juncture , however, we t ake di fferent paths : Chaney des cribes conceptua l plot s as mediating existence , that i s , standi ng between a human and the world ; I contend tha t those plots are the world . Let me try to make this di s tinct ion cleare r by dis- cussing how anthropo logist s have talked about cognition and semant ics . The great contribution of transformational gram- mar has been to come to grips wi th the problem of meani ng in l anguage , something descript i ve lingui st i cs i s unable to handle . But the t heory i s full of di ffi cult i es and at bottom rests upon a determinist ic posit ion of i nherent lingui st ic structures in the human brain . Cogni tion studi es have be come 1 61 enormously popular, but t hey too rest upon a view that sees culture as some sort of screen , s i eve , or map through which experience filters, giving onl y a distorted view of the worl d and s eparating us eternally from knowing the world "as it i s . " Since cognition and semant ic s deal primarily with lan- guage (we do field work by talking to informants-- we learn about their way of life to the extent that we can understand what they say to us) , what have those working in ethnographic semantics accomplished? Judging from Colby's ( 1 966) survey article , results in terms of "the semant ic codes of a partic- ular spee ch communi ty " are disappointingly slight , especially for tho se convinced of the s i gnificance of the actor ' s expla- nation-~which some have supplied with t he generic name of ethnoscience . Yet we do know a great deal about the lan- guages , values , concepts , and Gestalten of other cultures . Anthropologi s ts have been abl e to collect such data by lea rn- i ng the language , t alking with people , and participating in their lives . The difficulty, I think, hinges on the search for those codes : efforts are bent towards boiling down words and 1v0rd- making to first principl es from which alone can be generated statements that are linguistically or conceptually accept able to native speakers . This is determinism wi th a • vengeance . Unlike descri ptive lingui stics , "Ethnography still l acks anything comparable to the phonemic principle " (Col by 1966 :1 6), i . e. , a classificat i on of meta- meanings whose di stribution can account for all " semes " in a culture . 162 Colby ' s use of "lacks " and continued discussion in his arti - cle asswnes that the discovery of such a principl e is pos- sible , that meaning l ike sounds can be re const i tuted , like orange juice . Another reveali ng illustrat i on of the mental set of t hose searching for order in meaning is the work of Katz and Fodor (1 963) on semantic rules . Colby (1966:10) describe s their cont ribution as ... a semant ic metatheory describing constituents , ob jec- tives , and constraints of a ·workabl e semantic t heory . Accordi ng to t he authors, a semantic theory a ccounts , without recourse to the context ( e i the r lin uistic or non- lin uist ic for t he speaker ' s ability to interpret a sentence of his l anguage . A the ory which accounts for contextual i nfluences on t he i nterpretation of a n utter- ance would have to represent all the speaker ' s knowledge of the world , a requi rement which the authors consider unrealistic (italics added) . But , while dropping context because it s inclusi on would i npede the development of a generat i ve theory, Katz and Fodor state that a ... se~antic theory should a ccount for sentence i nterpr e- tation through (1) determi ni ng t he different readi ngs possible, ( 2) detect i ng semantic anomal ies , and ( 3) deciding on paraphrase re l at ions between sentence s (Colby 1966 :1 0) , goals that seem i mposs ible to even expl a in without refe r ence to context ! Dreyfus (n . d . ) argues that i t will be i mpossible to program computers to use language (o the r than programmed t ransl at ion "tricks " ) because language is context- dependent even for computers , and you car1not program the machine for all possi bilities . Exclusion of context , moreover , means that art i ficial i ntelligence i s "unrealist ic." 163 The art icle of fa i th sustaining ant hropologi sts , as oppo sed to lingui sts , interested in context ual factors in semant ic theory is that non-lingui st ic contexts are limited by commonality . Col by (1966 : 13) presents this view : Actually, there i s no reason to assume that the non- linguist ic context of speech, as conceptualized by the speakers, is so unique and rooted in part iculari ty of time a nd setting that it cannot be chara cterized rela- tively pars imonious l y . The anthropological faith that non- lingui st ic context need not be treated encyclopedic- ally--that somehow it can be made more simple--is reflected in frequent reference to "cognit ive structures " and related concepts (italics i n the original ). The results , e . g ., two classi cs , Frake 1 s 11 The Diagnosis of Disease Among the Subanum of Mi ndanao " and Conklin ' s "Hanunoo Color Categories , 11 a r e ob sessed with category , system, and structure , i mmobilizing meani ng and t hus human lif e i n plex- iglas s . On ce again t he search for order is tri umphant , every p i n and needle accounted for , the cultural audit complete . But all i s not well in cognitive theory . D' Andrade ( n . d . ) presents dat a that he thinks are evidence that s ocial scientists construct cultural reality falsely . Three sets of observations have been co l le cted of what D' Andrade assumes is a single event : lengthy discussion among groups of 3 under- graduates each, observed through a 1-way mi rror by an i nves- tigator . The first set of observations consists of 11 i mrnedi- ate recording" by the i nvest igator of the social behavior of each subject . The second co ns i sts of each subject' s rat i ng of the social behavior of the other 2 members of hi s group after the discussion event . The third consists of the 164 invest i ga tor ' s ratings from memory of the behavior of each s ub ject . The epistemologica l assumption behind the test was that i f the resul ts of any set of data di d not agre e with the othe r 2 , or if a l l 3 disagreed , t he data were therefore i nvalid . Wi t h these three different types of data i t is possible to ca rry out the same kind of anal ys i s of each type of data , and then to compare the resul ts . Such a comparison gives a general i ndication of the val idity of the te ch- ni ques , since t wo technique s which yie l d non- simil ar result s cannot be valid measures of the same thi ng (D tAndrade n . d .:1 5) . According to the findings , ratings of behavi or recorded during the event by the observer are diss i milar from e i ther set of rati ngs made by actors or observer after the event-- "long term memory ." D' Andrade conc l ude s that "memory dri ft " of how other people behave moves " in the di rect i on of the rater ' s conception of ' what i s like what '; that is , in the di re ction of the rater ' s construct ion of real ity ." D' Andr ade ' s conclusion i s t hat we are unable to pry culture off t he world , and that thi s i s a deficiency . 18 As a result of this type of memory dri ft , any attempt t o di s cover how human behavior i s organized i nto multi- behavior units , such as dimensions or clust ers , which is based on data consisting of judgments based on long term memory will result in co nclusions which reflect primarily the sub j e ct s ' construct ion of re~lity , not t he world as it i s (n . d .:1 8-19 ; italj_cs added ) . Therefore , re c onstruc t i on of social behavior (which i ncludes most ethnography ) i s inva lid . If t he argument presented here is corre ct then a l a r ge number of social s cience studi es are brought i nt o ~ues- tion . Studie s whi ch are based on corre l at ions fro m 165 memory based check- lists , rat i ngs , or i nterviews are obvi ously pl aced in the category of dubious f i ndin..gs . Studies i n which the co rre l at ions are based on quest ion- naire ~ i n which the re spondent answers on the bas i s of hi s recolle ctions also are placed in doubt ( n . d .: 20) . D' Andrade does not dis cuss so l ut ions or provi ae a measure of what const itutes long term memory . The logical i mplicat ions of his r easoning seem to be that (1) s i nce actors probably cannot i nteract and analyze their i nteractions simultaneously , any emic observat ions cannot tell u s what is "really happen- ing '' ; (2) any after- the-fact judgments of the observer 1 s are also invalj_d ; ( 3 ) the only way for the social s cientist to get at the "world as i t i s " is t o describe events as t hey happen , which implie s further that (4) an observer cannot be a part icipant or i n any way be present in t he event , other- Hi se t he i ndeterminancy principle comes i nt o pl ay ; and l ast ly ( 5) , i t would seem that the only way anthropologi sts can do their work is to watch people as George Scha ller watched lions--f r om behind a blind , or a 1- way mirror ; one certainly c ould not t alk to them . The culture here-world there view can be come ridiculous i n addition to being wrong . Generat i ve grammars , rewr ite rule s , s emantic theories of a ll sorts are trying to a chieve one thing : to re duce lan- guage and meani ng to a basic recipe wi th dire ctions for maki ng everythi ng else out of i t on t he assumption t hat l anguage acquis i t ion and use , and t herefore lingui st ic under- standing , pro ceed in ce rtai n orderly , rigi d , predictive ways . Whether it i s emi c orde r or , i n the long run , etic order , the 166 predication of sifting out generative/predictive order is probably based on the r eality that we can use language to fit correctly an infinity of s ituations--we do not speak randomly or ac cidentally . So, by extension, everyone is busily trying to nail those patterns down in what he understands to be a scientific manner . But despite Ctomsky ' s dismissal of an· "encyclopedic theory" (Colby 1966 :1 3), we . cannot understand a part wit hout the whole--in thi s case the sociocultural con- texts i n which language i s used--and if we c an never encom- pass t he whole , which bi ts do we i nclude and which do we l eave out? If you put back together the parts i so l ated in your analysis , have you reconstituted the whole itself? a simplified version? a primi tive version? something that has little resemblance to the original? When you peel l ayers off an onion , do you still have an onion? Can you put the layers back together again and restore your onion? So much for the mediat ion of experience . Yet there i ~ pattern , accumulated experience, and expe ctations . How are conceptua l plots the worl d? Perhaps a primitive analogy will help to understand the nature of the interstice . We have a picture puzzle of 600 pieces . I f one tri~d to put the puzzle together randor:1 ly , trying to match every projection to every notch , assuming 2 nubs and 2 notches per piece---and not adjusting for edge pieces--there is 1 chance j n 1200 of joining the first 2 pieces ; then 1 in 1198 for the 167 next match , 1 i n 1196 , 1 i n 11 94 , and so fort h . The puzzle would take a very long time to put together . The supre e challenge would be an all-white , circular puzzle . I t is a gimmick because it i s not an ordinary picture puzzle . But , disregarding gi mmicks , what about a pu_zzle wi th a picture? How to begin? We can sort the pi e ce s according to e dges and corners , color , or line s . Doing a puzzle without a helpfu l picture of the completed puzzle by your e l bow would be pos- sible , but it would stil l t ake a long time . To put the puzzle together according to the picture is to duplicate the picture . Without that helpful picture , can one generate the picture out of the s orted categories? Onl y superf i ciall y . For example, this puzzle has a lot of blue ( sky?) and green (trees?), and a small piJe of something red . Only when the piles are put toge t her does one realize that the blue is probably sky (see that bit of cloud) and the green i s a grove of trees . Yet that does not tell us much about the whole . There comes a point at which the puzzled person ceases to be puzzled and says , perhaps, "I' ve go t it !" or "Now i t ' s coming !n He sees what the completed picture will look like , and this seeing can occur with the fitting of a single pi e ce . The remai ning pi eces make sense through reference to the whole, which has becorne--bui l t up pie ce by pie ce but depend- ent upon relat ionships between pi eces-- the context . If the picture were of something out of t he ordinary , for i nstance blue trees and pi nk sky , the task would be more diff icult 168 because the puzzle would not refer to the real worl d . I f the puzzle were of a Jackson Pollock painting , t he moment of see- i ng wou l d be de l ayed for an even l onger t i me , but not so loz~ as with our gimmicky a ll-white puzzl e . We use l anguage in a roughly simil ar fashion, although to maintain an analogy be tween words and puzzle pi eces, we would have to limit analysi s of our l anguage user to an average middle-clas s American 2½ year old chil d , and yet that child could not possibly even begi n a picture puzzl e of 600 pieces . The po i nt about context--whose i ncl usion the current crop of c ognitive anthropol ogists di smiss as 11 unrealistic 11 -- i s the crucial one , a nd the sense i n which I am talking here a bout s emant ics and cognition . As J . Namrnour (1 973 ) has s hown , i t i s impos s ible to separate l anguage from t he world 1.-rn l earn to know and talk about through l anguage , and since language i s an essential part of culture , on a larger canvas we canno t separate culture from the world , or know t he world i n any way other t han i n culture . Puzzle piece s make sense only against t he l arger context that they i n the end form. Puzzle pi eces are the context . Words make sense only i n t he context of human experience . Language is experience . Non- linguj_st i c human behavior makes sense only i n the context of huma n experience . Culture i s experience . Violent human actions are inte lligible in the co ntext of human interest s . Reasons , intentions , and purposes are i nt erest s . 169 The s ignificance of such intere s ts may be becomi ng appa r ent t o some anthropologist s doing research on warfare , but they _still retain allegiance to the older order . By 1969 , Vayda had modifie d his eco - functional position toques- tion determini s m, stressi Lg the confusion of causal-effe ct with first cause, and admitt i ng the s i gnificance of human thought s and feelings , and of history . He counters White ' s culturological view of cultural evolution with a view that sees cultura l evolut ion as l es~ predictable , more random , and more fortuitous i n relationships between e nvironment and technology . What do these contrasting views imply for the i nterpre - tation of warfare among the Borneo tribes? Newcomb con- cluded his study of Plains warfare by saying that the "Plains cu ltures were warlike because they had to be , 11 [1 950 : 239 ] and we may suppose that I ban cul ture could , likewi se , i n a manner consis tent with White 1 s view of cultural evo lution , be i nterpreted as warlike because it had to be . No such i mputation of necessity or i nevita- bility is entailed by the other view of cultural evolu- tion Lwhich] ... allows us to note , where data are avail- abl e , the particular condit ions under which the I ban pa ttern of fighting arose and to note what we may call ... t he role of !! historical a cciQents " i n it s origi- nat i on . Thus we can cite evi dence on ~he recency of the custom of head-hunting among t he I bans ; the indications that the Ibans took i t over from other tribes among whi ch i t had been establishe d since an earlier time , a lthough in a different ritual context; anQ the i ndicat ions that the I bans were encouraged in head- hunting during the first pa rt of the nineteenth century by the Malays who made a pract ice of rewarding their Iban asso ciat es in piracy (prior to its suppre s s i on by Rajah James Brooke ) with the victims ' heads , a s well a s sometimes with a share of the plunder a nd sla ves .... What needs to be emphasized about these various circumstances ... i s that in t he early years of I ban head- hunting the thoughts and fee l ings of I ba n individua ls can be ar ued to have indeed ma tte r e d [Italics adcted .. If, as I think is the case , we cannot i nsis t that the a daptive effects conducive t o a 170 contiruation and spread of I ban warfare were operating from the outset , then whatever it was--whether Malay influence or something e l se-- that made the Ibans i n the beginning think that head-hunt ing was a worthwhile thing to do and thevery fact tha t they did think so may be i mportant . It can be argued that had they thought other- wise and had they not take n up the practice--and it is possible that such would have been the case under only very slightly different historical circumstances--then the l ater development of I ban warfare , which we have attributed i n part to its adapt ive effects , might have not taken pl ace (Vayda 1969 : 219) . Yet Vayda is cautious and denie s that his remarks are "an abjuring of determini sm ." Instead , he suggests tha t "there may be determir.ation without predetermination and that there may be orderliness i n cultural evolut ion ... and yet only a limi ted predictability" (1969 : 219- 20) . With special reference to I ban warfare , we can say that at those po i nts in it s evolution when it was still some- thing nev.r to t he I bans themselves the thoughts and feel- ings of the I bans about it s desirability may have been significant determinants of whether or not it would then become established among t hem (Vayda 1969 : 220 , ita lics added) . On the one hand , note that Vayda still couches the effects of tnoughts and feelings i n the l anguage of determini sm (and thus escapes the charge of uns cientific her~tic?) ; assumes that once an innovation i s no l onger new people do not have thoughts and feelings about its merit s any more ( some do , some do not) ; continues to think of human action i n terms of linear cause - and- effect . Note that he all ows the existence of order without predictability , an understanding of evol u- tion more consistent with the findings of biological evolu- tion , and that once warfare is accepted its effects can cause 1 7 1 tbe continuation of war and are thus adaptive : The circular chain of cause and effect is readily seen : warfare has effects conducive to the survival and increase of the I bans ; the survival and increase of the I bans are conducive to the maintenance and spread of Iban warfare (Vayda 1969 : 217) . This kind of ci rcularity i s not a tautology, unlike the Law of Cul tural Dominance . As I bans i ncrease reproduct i ve l y , t hey increase the number of people who share s i mil ar cust oms , beliefs , reasons , i ntent ions , and purposes . In the long run , however , Vayda g i ves more we i ght to the "adaptive effe cts " of warfare , e . g ., if our neighbors are warri ng against us, push- i ng us out of hunt ing and gardening lands, stealing our 'i,·,omen · and ki l l i ng our men , we respond by borrowi ng their war prac- tices and fighting back . I n so do i ng , we are no longer a peaceful people but a warring one . "A cause of the warfare by us may then be said t _o be the fa ct that the warfare by our enemies had adaptive effects for them" (Vayda 1969 : 218 ) . Thi s i s adaptat ion once removed . Although Vayda suggests that the I ban mi ght have chosen not to take up warfare , he does not explore possibl e alternat i ves . Cc nsequent ly, I per- ceive that the implicit adaptive advantage for us ( and adap- t ation is always seen as advantageous) is that if we do not borrow t he i nst itutions of our tormentors we wLll die out . While common sense may prompt us to such a response--fight back or die out--it may be our own cult ural belief that vio- lence stops violence . Thi s belief has been found erroneous time again , yet probably a majority of our population clings 172 lovingly to the deterrence hypothesis . For i nstance , pri or to the intervent i on of a British military peace - keep i ng f orce i n Northern I re l and, 80 peopl e had di ed i n 3 years of troub l es . Sin ce the arri val of the peace- keepers and their threat of vi olen t retali ation to a ct s of v i o l ence , nearly 8 00 people have died i n 4 years (Prasad 1973 ) . · While there are serious methodologi c a l problems in h i s studi es , Naroll (1966 , 1969 ) offers quant i tat i ve findings for both primit i ve and h i st oric warfare that at l east do not support the deterrence hypothes is . ( See Chapter II for further discussion o f Naroll' s work .) Although there is a very strong probability that i n the long r un the use of nonvi o l ent force will re sult i n far fewer casualties ( Deming n . d .), the attitude in our culture as expressed by one of my students who finds nonvio- lence unthi nkable is , "I would rather go down fighting !" Disregardi ng his i nsi s tence on adaptiv e functio n , Vayda is on to something that I cons i de r to be the most fruitful explanatory heuri st ic available at this time for understandi ng warfare of whatever t ype : a circula r chain . I have played with chains ; Harris expresses his pri nciple of technoenv i ronmental determini sm as a linear chai n ; except fo r its i nitial condit ion, the Harris - Divale the ory of pri mi- t i ve warfare i s a circul ar chain ; the Carne iro theory of the origin of the state posits a chain of events . Yet a ll of these are determinist ic: an element can have only 1 position in a sequence 1 and because it does i s i n the natura l order of 173 things a nd outside human volition . The economist Gunnar Myrdal talks about chains in a different way . There is pat - tern but not predetermine d order ; there is movement , but not along a line from here to infinity, or around in a flat circle, or in a simple feed back model , or towards a stable e quilibrium . Furthermore, unlike Vayda , Myrdal does abjure determinism and it s l aissez- faire , ex post facto consequents . What is wrong with the stable equilibrium assumpt ion as applied to social reality is the very idea that a s ocial process follows a direction--though it might move toward it in a circuitous way--toward a position whi ch in some sense or other can be described as a state of equilibrium between forces . Behind this i dea is another and still more basic assumption , namely, that a change will regu- larly call forth a reaction i n the system in the form of changes which on the who l e go i n the contrary direction to t he first change . The idea I want to expound . .. is that , on the contrary, i n the normal case there i s no such tendency towards automatic self-stabilization in the social system . The system i s by itself not moving toward any sort of balance between forces but is constantly on the move away from such a situation . In the normal case a change does not call forth countervailing changes but, i nstead , support- ing changes , which move the system in the same direction as the first change but much further . Because of such circul ar causation a social process tends to become cumu- lative and often to gather s~eed at an accelerating rate [e . g . , inflation, militarismj (Myrdal 1971 : 13 ) . Myrda l ' s notion of circular causation moves in a spiral , away from an equilibrium ; the effects of phe~omena can be to inten- sify themselves . 19 Let me borrow Myrdal 1 s illustration (1971 : 11-12) . Becaus e a man i s poor he cannot afford proper food ; becaus e he cannot eat well his health i s poor ; because his health is poor he cannot work ; because he cannot work he is poor . This chain , or vicious circle, of the circumstances 174 of poverty can never spiral upward out of poverty unless some one i ntrod Qce s a t lea st one new fac t or at one or more of several poss i ble points , for example, a guaranteed an nual i ncone or nationali zed health care . The task i s complica ted , however , because there is no pre dominant factor , e . g ., the nec onomic factor, " that one can readily j_dent ify and a djust by recipe to a l ter the direction of the spiral . Further- more , it is crucial to know how change in one fa ctor will aff ect the other factors before one begins to tinker (Myrdal 1971 : 19- 20) . Yet without the i ntroduction of viable changes , the chai n will spiral downward and i ntensify i nto Oscar Le1vi.s ' culture of poverty and the widening abyss in the mod- ern world between the have and have- not people s and nations . I n the latter i nstance the effect is compounded because weal th is concurrently generating an upward spiral . Or~ce a man has money to spare that , when put to work making more mouey by controlling tools and resources , can provide him wi th a living and a bit lef t over, he i.s no longer poor and i n fact becomes wealthier (Cf . Lundberg 1968 ) . Do the spi ra l s of cumul at ive causation have i nherent limits? A downward spiral does , i n the manifestation of I starvat i on and death . Any Malthusi an lower limit i s blurred , however , where modern medicine lowers the death rate through medi ca l trea tment of diseases that are symptomati c of star- vat ion . Any upper limit is even harder to define , but it exists . Myrdal (1 971 : 35- 36) gives as example s (1) the 175 a ccumul ation of old capital equipment that slows down produc- t i on and i s overtaken by developing i ndustrial areas usi ng new equi pment . England was di splaced by the Uni ted States and, l ater , Germany as the world ' s industrial leader l argely because of t h i s limit . ( 2 ) I f i ndustry and population be come too concentrat ed , t he burden of public expenditure may slow down or reverse the spiral of prosperity . The present condi- tion of American indus tria l centers such as Detroit and New York , where the i nner city has come clo se to bankruptcy and death , i s a mani festation of such a limit . ( 3 ) I f wages rise so high that t he product workers produce is no longer compet- i t i ve in a larger market , the spiral rapidly reverses . In t he pas t decade , products produced by t he highly pai d Ameri- can worker cannot compete wi t h l ess expensive products of higher quality made with cheaper foreign l abor . American factories close , and prosperit y is no more--for t he workers , that is . The upward spiral is s till in motion for multi- national co rporat ions that can readi ly move capital to wher- ever costs are lowest wi thout any risks to their prosperity . Thus the American worker i s on a downward spiral but American capital cont i nues upward carrying with it-- for t he t i me being--the Japanese worker . Although Myrdal 1 s hypothesi s is nearl y two decades old , it s address to the equilibri um economics of the 1950 1 s is not only s till relevant i n e conomics , i t i s a refre shi ng draught i p anthropology where so many are still "di s coveri ng" 176 equilibrium mode ls . I nclee d , Myrdal intended that hi s "prin- ciple of interlocking circ.:ular interdepender,ce within a pro- ce ss of cumulat ive causation " be valid for all social rela- tions (1971 : 23). I n a burst of irritation at Vayda ' s earlier (1968 ) attempt to explain primitive warfare as an equilibrat- ing response to disequilibrium between contenders , Aberle retorts : With respect to the topic at hand--warfare--it seems t o me , as a starting poi nt , valuable to assv_me that the a i m of much warfare ... i s not to equalize i nequalities, as is suggested in one of Professor Vayda 1 s hypotheses , but to i ncreRse them . Conceivably an equilibrium model might have some poss i bilities for application when the warfare ... occurs between groups at a similar level of organiza- tion a nd with s imilar technologies in similar environ- ments- - groups that are adjus ting to random fluctuations of population and/or resources . But when war occurs between groups with different technologies (military, productive , or both) or different levels of organization , or both , it i s typically an expansive operation for one of the systems in question ; it aims at a continuous dis - turbance of i nter-societal equilibrium .... In several parts of the gl obe we find very l arge contin- uous areas occupied by language families whose members have diverged from one another r e l atively recently in hu_man history . Tnis would be true ... for Bantu , Semitic, Inda- European and Sinitic . Many of these l in- guistic expansions car1 be shown to have swallowed up , shoved as i de , or pushed into refuge areas groups of dif- ferent linguistic background that earlier occupied t he area . These distributions suggest that expansi on of peopl es "on t he prod " is not a product of modern times or even of the historic period .... I suggest , then , that in terms of scientific yi eld we are likely to ge t further , faster with concepts like competi- tion , expansion, and domination , than with concepts like function , equilibrium , homeostasis , and reduction of i nequalit ies (1968 : 99- 100) . It i s not clear to me , however , that equilibrium theory is any more applicable to warfare be t ween similar societies than 177 it i s to dis s i milar ones . Why one type but not the other? Why either? We run the ri sk of be i ng caught i n t he futile t rap of adapt ive versus maladapt i ve effect s : if one breaks even or wins , warfare is des cr i bed as adapt i ve ; i f one los e s , i t is maladapt ive . Harri s (1 97 1) i dent i f ie s Yanomamo war- fare , the pr imary or t ype case for the Harri s - Diva le hypothe- si s , as maladaptive , and the se pe ople fight only each other ! Lastly , couldn ' t one a l s o descri be warfare between diss i milar s ociet ie s a s a n adjustment " to r andom fluctuations of popula- t ion and/or resources "? Whe re i s the pl ace of human volition i n Myrdal rs spiral? Can the poor man choose to be healthy? Can he will foo d i nto existe nce? In t he period of t ransi tion between the capita lis t and we l fare states , the good libera l winces a t conservative charges that the poor are l azy and no-goo d and i t i s all their own fault . The goo d liberal goes to consid- erabl e l engths to sho,v t hat the poor man is a victim of cir- cums t ance s beyond his control . Yet, while the poor may be vict i ms , they a re not blind or completely helpless . They can , and have,protest and revolt aga i ns t tho s e who perpetuate economic i nequal it ies, and their rebellion need not be vio- l ent . The poor man is poor i n a larger context , one tha t i ncludes t he rich man . Eve n for those who are born i nto wealthy familie s , we do not a rgue that the man ' s wealth i s beyond hi s control, that he i s a victim of cir cwnstance s , that ne ca:u.not stop his money from maki ng more money . 178 Volition is probably the most i mportant factor i n being rich--otherwise you mi ght become unrich very quickly . At the heart of all welfare legi s lation and proje cts to eradicate underdevelopment and i nequality i s the necessity f or the rich to choose to help the poor , to choose to redistribute income and a ccess to resources . The state , in Myrdal ' s estimat ion , has always owed some measure of i ts existence to the popul ar will (1971 : 44-48 ). Even at least some of the great reigns of terror persisted because of popular support , based upon a variety of reasons, e . g ., Tl acaelel and Moctezuma , Shaka , t he French Revolution , Hitler . Those few modern states that have achieved an egalitarian economy based upon redi stribu- tion have i nterfered with the spirals and introduced new factors because of the wills , wishes , idea l s , and reas ons of t hat segment of the population that had the power to do so , thus i ntensifying and spreading egalitarianism, and recre- ating the most ancient form of human society . Although equili brium theory does not appear to work in economics or anthropology , it is part of the cultural context of both our economic act i vity and our warfare . The whole arms race , for i nstance , i s predicated on t he assumption that i f an equilibrium in armaments is achieved , war will not occur . This assumpt i on is a corollary of the deterrence hypothesis . But , speaki ng eti cally , balance of power poli- tics has a ca tch to it : a bala nce exists when we have the edge . Since all other contenders hold a s i milar be lief , 179 there is no stabl e equilibrium but an ac celerat i ng spiral of cumulat i ve causat ion that already has blown ma ny people to kingdom come . So much for equilibrium between societies of similar technologies and organization . It is what the actors i n our time believe and want that powers the spiral , and n o natural l aw of equili brium can possibl y be come manifest . I n trans ferring the heurist ic fr om economics to warfare--and it i s intende d to be transferable--the spiral has a vital char- acteristic . Wars and events leading up to them seem to possess a dynamism that few other so ciocultura l phenomena have . There is human volition , de cis i on , and implementation . Even when we speak of events taking on a momentum of their 01:m at some point , a closer look may reveal a human choice s ·1..ch as an ult i matum . Ultima tums by their nature and defi- nit ion set a point of n o r eturn . Yet it is human not ions of bluff , or honor, or i mpatience , or manipulation of events to provoke an excuse that direct the issue of an ultimatum and a response to it s acceptance or rejection . The spiral of actions and reactions sometimes seems to tighten and i ncrease i n momentum--the I ndo china W~r is a good example-- unless another factor is i ntroduced to alter the spiral, such as a change i n foreign and military policy . And human beings make such policy . Thi s is the end of my critical examina tion of current t he or ie s prevalent in anthropologica l thinking that are determini st and unitary i n desi gn . I have tried to show what 180 some of the consequence s are of philosophi ca l dete r minism and of unitary theory i n socia l s cience . I n particul ar , I have argued that upon close i nve s tigation the theories not onl y compete wi th each other but , when taken s i ngly , appear to create more probl ems of understandi ng than they s olve . I have concluded that the qua lities of orde r assumed by philo- sophical determinism are i ncongruous with t he huma n a ctions they are meant to expl ain . Since warfare occurs in a variety of cultural patterns t hat are not nece ssarily or sufficiently predictabl e , understandi ng human action recogni zed as war must be in terms of the context in which it occurs and from the human beings who experience it . In the end , as psycho- logica l and functional explanat ions prove i nadequa te, one must re ly on geography and h i story to t ra ce out the cultura l patterns . I n the second ha lf of this the s i s , I shall try to re concile t he Trukese to the European ethnographer i n a quan- titative study that i n cludes as variables a few human reas ons , i ntent i ons , and purposes . As Berreman hopefully declares I would prefer that we Trukese ethnographers not give up our Tru.kese methodology , but that instead we define , explicate and the1eby i mprove it . For I think it i s possible t o demonstrate that it works- - that i t ge ts us t he re-- when well and properly done . What 1,-;re have tried to do i s fi nd out how one goes about do i ng i t well and properl y . And there may be many di fferent ways of do i ng it ; or a few princi pl es which can be applied i n diffe rent ways , or many techniques which work i n various c ombi - na·t ions . Perhaps s ome new t echniques need to be added-- Trukese navi gators , after a ll , have by now a cquired 181 coKpasse s for use i n eme r gencies . Eur opean ethnographers , meanwhile , mi ght unbend a bi t - - try c omi ng out of t h e chartroom and ob serve more of the worl d around t hem . Each appr oa ch has s omethi ng t o s ay to the other . Ti me will tell wha t i s us efu l a nd wha t i s no t (1 972 : 231 ) . Yet I woul d rathe r re l y on s omething othe r than the judgment of h i story to a ssay whethe r or not one i s wasting one ' s t i me . As it i s not wi se to declare a pri ori what t he lives of huma ns are about , ne i ther i s i t just to trust only a nosteri ori . 182 CHAPTER II METHODOLOGY The Logic of Cross-Cultural Studies 1. What are we looking for? I n the first chapter I discussed some of the salient theoretical issues that directly and indirectly influence the substantive outcome of warfare studies . The next consider- ation is of the re search methodology I have used i n the pres- ent study. The primary value of the cross-cultura l comparative method i s to provide empirical testing of an hypothesi s on a large number of cases with findings presented i n a mathemat - ica l statement . The determination of "large " is difficult , i n addit ion to the more obvious problem of the i ndependence of cases ( the Tylor-Galton problem) . There is a variety of other problems in cross-cultural research that may call the method ' s utility into question . 2O Anthropologists who con- s i der themse lves humanists find the method rude , crude , and dehumanizing . The new wave of scientism finds it necessary but not sufficient and deplores the inability of synchronic comparisons to produce cause-and- effect statements : Statistical cross- cultura l surveys can , indeed must, be used t o supplement other modes of generating and te st ing 183 hypotheses , but they cannot be use d a l one or even as the primary sources of nomothetic statements (Harri s 1968 : 618) . The source fo r such statements i s 11 ••• detailed diachronic and synchronic causal- functio nal anal ysis of spe cific cases " (Harr is 1968 : 633 ), that i s , the analyt ic and creative powers of i ndividual i ntelligences . There seem to be jus t three ways to go . (1) An indi- vidual may make a universal generalizat i on about human behav- i or t hrougn analysi s of one or a few cases . Thi s falla cious hast y generalization in the name of science is highly disap- proved of and has been labeled case illustration (McEwen 1963 ) . It a l so suffers from circul arity : the illustration is used as proof of the t heory . An ad absurdum exercise i n thi s metLod is Lionel Tiger ' s a cceptance of William Golding ' s fict i onal illus trat ion of human bestia lity as proof for bio- l ogi ca l determinism and i nnate depravity . Less exciting exampl es are Dival e ' s (1 97 1) use of 6 societie s upon whi ch to base a theory of warfare applicable to all pr i mitive soci- eties , and Colin Turnbull ' s (1973) a c count of the dreadful Ik , a disrupt ed cult ure t urned parasitic and can cerous , f or which Turnbull re commends radical surgery before the social di sease spr eads and corrupts other societie s . Hi s conclu- sions will delight Ayn Rand and her followers , who will prob- ably parade the I k case as evi dence tha t humans are by nature pri marily selfi sh and that subsequently her theory of objectivi sm i s a fa ct . 184 (2) Or one can put a l arge nuJnber of s ociet i es coded for a variety of vari ab l es i nto a comput er and s e e what vari - ables come out sticki ng to each other . Harris (1 968 : 632 ) sui tably but di sapprovingl y calls thi s " the-throw- i t - agai nst - the- wall-and- see - if-it - st icks " te chni que . Ye t such pure i ndu ct ion i s logically i mpossible , s i nce f a ct s do not speak for t hemse l ves (Hanson 1958 ; Chaney 1973), and dat a a r e i nvari abl y coded i n a ccordance with a t leas t t a cit a s sump- t i ons . The cross-cultural comparisons cont a ined i n this dis- sertat i on appear to be of the throw- i t - agai nst-the-wall type , but as dis cusse d i n de t a il be l ow t he se l e ct ion of v a riables and the i r cod i ng have been done i n a ccordance with a t heory , albe i t amorphous , of human a ct i on . ( 3) Or one can do as Harri s advi ses , which i s i niti- ally s i milar to (1) above : Derive a generalization/ hypothe- sis t hrough anal ys i s of a few cases (Hans on ' s r e tro duct i on ); look for ant e cedent variables in a l arge collect ion of cases ; put the data i nt o t he computer a nd see what variabl es come out n ot s ticking to each ot her (as so c i ation) but precedi ng and following each other (cause and eff e ct ). · I f we wish t o achi eve an orde r l y picture of h i s tory , we must begi n by assumi ng t hat there are certai n orderly principl es which are at work . And we must u se t he se principles to order and classify the data .... we must process the data , question i t , classi fy i t , and code i t i n re l at i on to the expecta t i ons of our ma j or premi ses . Then and only then can we a ccept a f ai l ure t o s tick to the wall as evidence a ga i ns t the major premise . At the same t i me , it i s onl y from such a procedure t hat the par- ti cul ar corre l at i on on each tri a l may be seen as 185 connecte and governed by the general princi pl es of sociocultural evo l ut i on (Harri s 1968 : 632- 33) . Those pri nc i ples , of course , are to be those of cul tural evo- l ution and cultural materialist causality . Beneath the appearance of s cient i fic rigor , I thi nk one shall find the nemesis of unfalsifiabi l i ty . On t he one hand , how c ould t he major pr emi ses be f ound i n error i f the mi nor premi ses ( data ) are se l ected a ccordi ng to the ma jor premises ? On t he other hand , an argument can be valid although t he premise s are false as l ong as there i s internal consi stency . The point i n quest i on is the truth or fa l s i ty of those premises--and we have the right t o inqui re after truth s i nce t hese macro- theories purport to a c count f or as pr ofound a mat t er as order in history . I f I understand Harris corre ct l y , while he a llows for the possibi l ity that properly done ( styl e 3) stud- i es c ould admit di sproof , exceptions to regul ari t ie s , and unsuspe cted relationshi ps , he seems t o view t he absence of relationshi ps as indicat i ng that the research i s not a ny good rather t han entertai.n the possi bi l i ty that there is n o pat- tern or regularity of the sort he assumes to be t he case . Let us admit that posit i ve correlat i ons established in the face of the hazards of codi ng and the unre l iabilit y of the sources may nonetheless , barri ng systematic bi as , merit our conf i dence .... There remains , however , the probl em of the no ncorrelat i ons which may arise from poor ethnography , biased analysi s , faulty c odi ng , and i mproperly phrased hypothesis (Harris 1968 : 632 ) . Thi s expl anation of noncorrelation i s identica l with that of Kobben ( 196 7) , who i nsists that where exceptions to a 186 correlation occur, it i s not due to the nature of the excep- ticn but to faults such as those Harris lists . While I cer- t a inly do not deny that human error may be a significant explana tion for some exceptions-- much cros s-cultural work i s appallingly sloppy--we cannot assume that i t must account fo r all exceptions . The presumption of ubiqui tc,us order i s patently a pri ori. Chaney ( 1972) has dealt wit h the possi- bib.ty of little or no regul ari ty and , unlike almost a ll anthropologists who have written on philosophical i ssues , does not find such absence threatening : ... it behooves soci al scientists to realize that s ince the degree and nature of regulari ty in sociocultural phenomena is still an open qu~st ion , to state t hat no regul arit i es have been found is just as i mportant as to find regularities . I f one does not know one way or the other , it is just as significant to i mplicate the ext ent of no regularities as to explicate the extent of regu- larities (1 972 :1 3). Taki ng users of the comparative method as a whole, we are l ooking for 2 things : the presence and absence of joint occurrences of variables and expl anations for the origin and perpetuation of these presences or absences as due t o e i ther fortuitous circumstances (geograpby and hi story ) or natural order (psychology and funct i onalism) . Research workers i n ar..thropology heavily emphasi ze the presence of joint occur- rences and prefer to explain their existence in psycho- functional terms . Thi s emphasi s has been traced , i l l us - trated , and discussed throughout this paper . Even when his- tory i s reluctantly admit ted as significant , the natural 187 order of things i s i nvoked to expl a i n the adopt i on and per- petuation of a known tra i t . When Dri ver (1966) explai ns the di s ribution of ki n avoidance behaviors i n terms of ge ography and hi story , s ome commentators admi t the possibility of dif- fusion t o a ccount for t he or i gi n of a cultural t rai t but assume t hat it s perpetuation must be due to psycho- funct ional f a ct ors , i . e . , .. . t nat t he i nformat i on in quest i on (a ) me ets certain soci o- psy chologica l needs and/ or ( b ) func t ions to mai n- tai n some vari ab l e cr i t ica l to t he surviva l and we l l - be i ng of the p opul at ion (Col lins 1966 : 149 ) . Therefore , what exi sts must exi st of necessity . One must remark that such fa cile explanat ions are post h oc and i mpl y a belief i n t he natural benevolence of so cial i nst i tut i ons ; neither explai ns s ituat ions i n society whe r e human be i ngs are de pri ved and mi serabie . Another r epl y t o Dr i ver a r gue s t hat ... the vari abl es tradi t ionally asso ci a te d wi th avo i dance of ki n are he re shown to be l ess i mportant fun ctional corre l ates t han cu l t ure area and l anguage . I f thi s be true ; we a re s t i l l l ef t wi t h t he quest i on , why? Dif fu- s i on i s no answer . Most ant~ropologi s t s s i nce the t i me of Patterns of Culture have agre ed t hat a t rait i s "a ccepted " if i t "f i t - in- with " t .t1e other ma jor e l ements of the 11 taki ng " cul ture . The f inal expl anat i on must therefore be a f unct i onal one (Freilich 1966 : 153 ) . ~-'hus we have anthropol ogy by consensus and a c ceptance entai l- ing purpose . The assumpt i on of f un c t i on i s perhaps as deepl y imbedde d in the l anguage of anthropo l ogy as i s adaptat i on , although the case f or it s ne cessi t y i s by no means as c l ear . Yet even Driver assumes some ki nd of psycho- funct ional mot i - vation , although it may be a ne cessary but not sufficient 188 condi t i on for the diffusio11 of a trait : Pro bably all the psycno-functional "causes " of kin avoia- ances advo cated by Tylor , Frazer , Freud , Lowie , Murdock , and Stephens and D' Andrade have had some i nf luence on the or igin , mainter..ance , and di spersal of these be haviors . Even the most extreme ge ographical- historical enthusiast needs a package of psycho-functional "causes " to get the avoi dance behavior started . Once such behavior has become firmly establi shed , however, it seems to diffuse by intert r i bal marriage to peoples who l a ck some or even most of the "causes " discovered so far . It also fails to occur among some peopl es who possess most of the "causes " (Driver 1966 : 147) . But what is going on in that package i s s i mply not clear . Barnet t (1953) manages not to deve lop any such package in his e legant and fine l y detai le d work on the origin and acceptance of i nnovat i ons , in which he argues that needs are as likely t o be generated as satisfied by innovations ; that well- being i s meani ngful only to the i ndi vidual; t hat people , not cul- tures or societies , accept or reject innovations in the light of many cons i derations ; and that what "fits-in- with" the life a nd expectations of one i ndividual may be anathema to her neighbor . We do know that patterns exist , t 11at human social life and cultural experience are not random or chaotic , although t 11ey are very untidy (Dougl as 1966) . For i nstance Driver and Massey (1 957) have mad.ea strong case that change in gross sociocultural phenomena tends to progress i n acer- t ain sequence : division of l abor , residence , land tenure , descent , and cousin te rminology . Eggan (1966) , a pparently independently and from a different t he oretical position , recons tructs a similar progression in change among several 189 groups of rorth American I ndians . Nevertheless, I am l arge l y i n agreeme nt with Chaney , who argues t ha t "the ' patterns ' of sociocul tural phenomena are best conceived of as holding in terms of the cor.t i nuity of.§: culture ... rather than in terms of something i nherent in sociocultural phenomena analogous to ' cosmic glue '" (1973 : 1370) . Thus , we must deal with specific groups of people and the specific behaviors and beliefs that people who live together share . Even at this l evel one must be careful not to make the patterns more homogeneous than they are . In the end, the only point we can be clear about is that the tables , numbers , and coefficients cannot reveal the truth : " ... statistical manipulation of sociocultural data can only summarize the data . It can not yield di rect i nfor- mation as to functional and/or causal factors " (Chaney 1973 : 1368) . What statistical manipulation can show i s how often variables that I select seem to occur together, and whether or not there is variation i n this frequency from one geo- graphical region to another . It is entirely up to me to make the ass ociations and their variability intelligible. 2 . How is it done? The bi ggest i nitial stumbling block and source of error in con~arat ive work is the famous Tylor-Galton problem : how to be cert a i n that each member of the sample is an i nde- pendent case and not a duplicat ion, t hrough diffusion , of 190 another , which would i nflat e t he strengths of the associ- at i ons . I t i s also referred t o as "controll.i ng for hi story , 11 i . e . , eliminating those cases i n which an i nst itut ion or behavi or i s borrowed r ather tnan originat i ng i ndi genously . Here i s an apparent contradi ction . One is conce rned with the l aws of history but not with "mere 11 history . Laws of hi story shoul d be transl ated i nt o psycho-functional relat i onships . Cons i d.erab l e ef for t has been expended i n i nve nting ways to draw a pristine sample through statistics r ather than rules of thumb . Naroll (1961 ; 1964 ), Naroll and DtAndrade ( 1963 ), and Driver and Chaney (1 970) have i nvent ed various s ophisti- cated te chni ques , 4 of which Murdock and White (1 969 ) used to deve lop their Standard Cross-Cultura l Sample , the one I have ~sed here . This Sample , cu l mi nat ing over a de cade of work , nas been developed so that re search workers devoting energies t o any problem would turn out comparabl e findings . Prior to t his , everyone developed his own sample wit h highl y variabl e standards and no confi dence t hat the sampl e at hand was com- parabl e to any othe r . Apparently oblivious to Murdo ck 1 s effort , Naroll (1 970) has developed hi s own standard sample , and eventually we shall be faced with a choice between the 2 and a revi val of the cont i nui ty probl em . I n a ny event, Murdock and White have carefully pinpoi nted t hi s ocietie s i n their sampl e a s to exact geographical l ocation and specif ic date to which the l argest body of ethnographic data pertai ns . To s ol ve t he Tylor-Ga lton probl em and allow for geographica l 1 91 strat i f ication , Murdock and White have gone through one l as st ep i n classifying cultures according to similarity , pro- ceeding beyond clusters (Murdo ck 1966 , 1967) and sampli ng prov i nce s (1 968 ) to sele c ting 1 society from a l most a ll of t he 200 sampling pr ovinces . Gi ven the demands of pinpoint i ng 2 provinces are unrepresented ; 2 are fur t her di v ide d i nt o 2 again and a repre sentative cho s en from each of the 4 ; 14 are unrepresented because f urther judgment has been made that they shou ld be l umped without other provinces . The revi s ed pr ov inces or "di st i nc t i ve worl d areas " number 186 , 1 s ocie t y r epr esent i ng each . Of the 186 societies , 141 have food- produci ng ecoEomies ( 56 havi ng intensive agriculture , 19 hor- ticulture , 51 swidden ; 15 pastoral) and 45 foo d-collecting economies (13 primarily gatheri ng , 14 hunt i ng , 18 f i shi ng) . Politically · and geographically, the sample i s strat i f ied as fol l ows (Murdock a nd White 1969 : 340 ) . 192 Major legion Stateless Minimal Smal l Large Total Societies States States States Sub- Saharan Afri ca 8 9 6 5 28 [27] Circum- Me diterranean 8 5 1 4 28 [ 29] East Eurasia 1 1 10 2 1 1 34 Insular Pacific 1 3 1 1 6 1 31 North America 23 7 2 33 South America 23 5 2 2 32 Total 79 50 23 34 186 The pinpointed time periods have the foll ow i ng di stribut i on (Murdock and White 1969 : 34 1 ) . Time Period A C-M EE I P NA SA To t al 1750- 1 B. C. 0 2 [ 3] 0 0 0 0 2 [ 3 ] A. D. 1- 1500 0 0 0 0 2 A. D. 1501 - 1600 0 0 0 0 2 [ 3 ] 3 [ 4 ] A. D. 1601 - 1700 0 0 0 0 2 3 A. D. 1701 -1 800 0 0 0 3 5 A. D. 1801 - 1 900 9 7 1 1 4 1 6 5 52 A. D. 1901-1 950 1 9 1 3 1 7 [ 1 6] 1 6 7 1 9 [ 18 ] 91 [ 89 ] A. D. 1950-1 965 0 4 [3] 5 [ 6] 6 0 2 1 7 Total 28 [27] 28 [29 ] 34 31 33 32 186 De viations from the Standard Sample in this study with regard to these cl assifications are indicated in brackets . Expl nat ions for such deviations are given below . 193 Murdock and White desi gned t he Standard Sample so that a re earch worker ne ed not use all 186 s ociet i es but could i nclude every se cond or third soci e ty as 11 ••• either being equa lly representa tive of the world ' s known and vre ll descri bed cultures as well as exhibiting slightly less con- tami nat i on from hi s torical influences " (196 9 : 352) . I de cided agai nst thi s option be cause of the uneven quality of war far e data : any way I chose , too many of the best - do cumented soci- et i e s were l eft out . Furthermore , for statistica l validity I consider any fewer than 100 cases to be unreljable , and f or the corr elations within each geographical area , I do not want t he J.irs to drop be low 20 . Last l y , one of the contri but ions of t his study is to compare the results done on smal l samples ·with f i ndi ngs from a large s ample . While I have worked wi t h the entire 186 societies , because of the unevenne ss of the data I have to exercise a no ther option stipulated by Murdo ck and Whi te , and that i s 11 ••• sub s t i tute other so ciet i es from t he same di stinctive areas or sampling provi nces without sacrificing any of the advantages except possibly that of i nte rcorrelation with the results of other s tudie s " (1 969 : 352) . I bave noted the 10 subst i tutions as they occur in the list of Standard Sample societies (Appendix A) . Sample size as a serious co nsi deratiori has been gene r - ally avoided i n anthropol ogy and relegated to l ast place after problems i n sample selection , bias , coding , a nd the e thnographies ( e . g . , Robner and Pe l to 1970) . Indeed , in 194 response to Chaney and Ruiz Revilla ' s (1969) findings that small saraples of 48 , 60 , 110, and 112 soc i eties produce cor- relations that can be significantly different from those based upon large samples, there may be a minor movement afoot to just i fy earlier work based on such small non- stratified samples with the claim that tney are mathemati cally superior to l arge strat ified sample s ! Rohner and Pelto , reviving an argument by Benfer (1 968 : 950) that 11 ••• all things be i ng equal , t he larger the sample size , the greater the probabil- i ty of the part icular test rej ecting the [ null hypothesis ] 11 ••• and therefore anthropologists should use small samples , without botheri ng to give any indication of what they mean by "large " and "small ," and bolstered by Meehl ' s (1 967 ) findings t hat any intercorrelation usj_ng a sample of 55 , 000 will be significant , argue t hat "re searchers need not hesitate to us e sa~ples as srnali as forty- eight or s i xty societies " (1 970 : 1455 ) . Yet Rohner and Pelto are not consi stent in t heir recommendations . After dismiss i ng Chaney and Ruiz Revilla t s findings (1 969 ) as not proving anything be cause the 7 sampl es were drawn by di fferent procedures , which is just the effect that Chaney and Ruiz Revilla were testing (Chaney 1970 : 1458) , attribut i ng vari ation among sampl es as due to 11 ••• a vari ety of extraneous factors rather than sample size itself" (1970 : 1454) but not speci fying what t hese factors might be , and JI ur ging researchers to go ahead and use sample s as smal l as 48 , in the end they recommend samples of "at least " 100 to 195 all ow for contamination and interven i ng variables . "I f one ' s cample contains only 50 or 60 cases . .. subsample operations may be come meaningless be cause of the smal l number of cases involved " (1 970 : 1455) . The ultimate j ust i fication for using small samples i s that t he user " . . . is sel dom in danger of erroneously drawi ng positive conclusi ons regarding the asso- ciation bet ween variabl es" (1 970 : 1455) , that is , Rohner and Pelto a re cla iming , on the basi s of Chaney and Ruiz Revilla t s work, that if one come s up with a signi ficant correla t i on us i ng a sample of 48 , one can assume t hat the correla tion would be even stronger using ever l a r ger samples . They do hint at the possi bility of samples be i ng too small. Si milar reasoni ng seems to be u sed by Divale (1973 ), who argues that U3 i ng data from different time periods and different c ommuni- ties of one s ociety--in contrast to Murdock and White ' s (1 969 ) i nsist ence on using data only from a pinpointed t ime a nd c om- munity--i s perfect l y all right because su ch di stort ion pro- duce s only r andom error which lowers correlations . "Thus , if significant associations occur , t he actual correlation is even higher than the one found , and the danger of spurious correlations can be i gnored '' (Divale 1973 :46) . Followi ng the recommendations of these gentlemen , we woul d be exercising sci ent ific vi rtue by us:Lng small sampl es whose data is taken from any old time or pl a ce identified as "Cheyenne " or "Maori . " The seriousness of sampl e s i ze becomes even more i mportant when one observes that many cross-cul tural workers 196 either claim or assume that their f i nd i ngs apply not just to the sample from which they are deri ved but to the uni verse of all human cultures . The most di ffi cul t aspect of quantitative research i s coding : maki ng hundreds of decisions as to how to classify gross data . One i s constant l y pl agued by doubt about one rs judgment. Since coders are actually further classifying the ethnographer rs classifi cations , error can be compounded . Often the very possibility of cross- cultural quantitative research is questioned because of the numerous possible and actual i nstances of error . Why even begin i f the ethnogra- phies are not reliable? Such reasoning has probabl y stopped many potent i a l research workers cold . The only a l ternative , if one i s g oing to do research in anthropology , is t o go out and be come another ethnographer ( but a reliable one, of ccurse) . Rechecking f i eld a ccounts i s a rare prac t ice, f or some obvious reasons--the people are deau , the cult ure i s gone , it i s thought better to go where no other anthropol o- gist has gone before--and one not so obvious . The rare restudies, for i nstance Oscar Lewis ' work i n Tepozt l~n fol- lowing t hat of Robert Redfield , can be in agreement or not . Do Lewis ' findings , which frequent ly are at odds with tho se of Redfield , mean that Redfield ' s descriptions are wrong , that he did not see correctly? Or that Lewis is wrong? If ethnologi st s are expected to analyze their own deocriptions (field data) and such work is a cceptable , why 197 then scorn the sources of the comparative research i:-rorker? Muc· of the goodness of etnnographie s must be taken on faith--inde e d , we do so all t ne t i me when we teach--but tnere are other philosophical justifica tions for doing just t hat . Cody (1 967 ) coLt ends that the d~ st i nction between description (what t he ethnographer observes) and action (what t he people observed are doing ) is a f a l s e one . For t hose of us t rained so assiduously i n relativi sm , such a thought i s nearly unthi nkable . For i nstance , i n examining t he classic witness expe riment--someone rushe s into a classroom, assault s the teacher , l eave s , and 20 astoni shed students are then a sked to wri te down what they saw--we can make 1 of the foll owing con- clus ions as to why no 2 of the descri ptions are identical . So e are in error (t he students di d not r eally see what hap- pened ) and some are not , but whi ch one s? I f we take all the points i n common to the 20 des criptions , we might have an a ccount of what actuall y happened , a composite or summary . Or we coul u say t hat t here i s no way of even tal ki ng i ntelli- gi bly about what really happened , be caus e our only evidence of what really happened i s i n the 20 aiffer-ent des cript ions . That is, l ogically and really , there are 20 di fferent events . Yet, bowi ng to pressures fo r conformity i n the data , more t han one coder can go over the et lmographi es i ndepend- ently , t he co ding de ci sions compared , and discrepancies accounted for i n mo re traditiona l ways , perhaps i n most i nstances by havi ng more c oders g o over the same data until 198 some congruence is achieved . Po~sibly t he only real benefit of reliability coding i s to spread around the responsibility for de cision~maki ng . It makes the research worker feel better . There is the continual temptaticn to code in accord- ance with i mplicit or explicit hypotheses , and to i nfer care- lessly . What is considered to be adequate codi ng? The purists among cross- cultural researchers tend to throw out any case in which the trait in question is not specifically reported as either present or absent . Others have devised ways (cften quite i ngenious ) to get around this problem . A frequent solution has been to examine each case within a l arger whole--either the con- text of the entire ethnogr aphy or the tot a l ity of what the researcher already knows of t he society or of the culture area of which it is a part . I f in the light of this contextual examination the re searcher concludes that a given· trait i s either present or absent , regardless of whether t he ethnographer specifically says so , he will code it as such-- normally using a special device to j_ndi- cate a lesse r degree of reliability . I n addition , a con- tent analysis of the ethnographic text may reveal that the ethnographer carefully and fully covere d the context , e . g ., puberty and religion , in which the trait , e . g ., circumcision r i tes , would be expe cted to occur ; in the absence of any mention of circumcision rites , the rater may feel just ified in rat i ng the trait as absent for that particular society (Le bar 1970 : 716) . The latter technique runs the great risk , I think, of desi gn- i ng the data to fit the hypotheses . I have used infe rence sparingly . I have found myself confident of my own co ding only through reading entire ethnographies even when the eth- nographers are explicit . I am uncomfortable with t he Human Relations Area Files' sorting of warfare data and find it necessary to put a culture back together before taking it apart . Lehar (1 970 : 716) suggests an acceptable reliability figure of 80 per cent agreement betwee n any 2 coders . The 199 le ss "objective " and the more " judgmental" variables a r e , of course , the lower the re l iability f i gure will be . I n this study , for i nstance , there are 2 attitudinal warfare vari- ables that could have profited from better operat i onali za- tion . One , however , was not used in quant i f icat ion , and agreement among coders for the other was very c l os e . Af t er all t he tedi ous work of sampl e se l e ction and coding , and the perenni a l frustrat ion i n warf are research of s canty data , one reaches the point of i t a ll--maki ng those mathematical s t a t ements . Even though there are r elat i vely fe w stat i stics i nvented f or nomi nal data as compared t o ordi- nal dat a , re sea rch workers i n anthropo l ogy seem t o be partial only to phi ( and a re l at i ve ,~ point-biseri a l) and chi - square , respective l y a measure of t wo- way associat ion bet ween 2 dichotomized variabl es and a measure of such an a ssoci- ation ' s significance . Somet i mes onl y chi- square a nd a proba- bility value are gi ven . Beyond dichotomous subcla s s i f ica tion of di chotomized variabl es and performance of phi and chi - square upon them , and an occasi onal scalogram , there i s n o fur ther statist ical mani pulation by and l arge i n quantitat i ve warfare research . The greatest amount of effor t i n such studi es i s expended i n i nt erpretation of the coefficients , explaining t heir relationships i n causal terms . Consis tent l y , coeffi- cients of association are not treated as summaries of the data , as Chaney defines their utility , but as l icenses to 200 build cha i ns of cause and effect to stand universally for a ll cultures throughout time and space . Refinements of this i nterpre t ive manner are demonstrated by t hos e research work- ers t hi nki ng i n the evolutionary paradigm who arrange the pai rs of a ssociat ions in a deve lopmental sequence, also to stand universally . In the following . section , which deals with specific authors and their contributions to quant itative warfare rese arch , I shall look closely at individual s tudie s in terms of me thodology, hypotheses, and i nterpre tation of resul ts . 3 . Who has found what? I n this section I discuss in var ying detail the con- tributions of 8 a uthors to quant itative warfare studies in anthropology . I t is a nearly exhaus tive review of what has been offered as significant s t at istical statements about the nature of warfare . Such a de tailed review io necessary not only to establish the prece dents for my own work but to fil l i n. the background against which my researc_h ca n be better judged . I deal with the authors i n a certa in order , pri mar - ily t he date of their work and secondarily vague themes in or i entation , from general omni-vari abl e studie s (Wright , Broch and Galtung) to problem-or iented studie s with particu- l ar points of view : political (Naroll, Otte rbein, Ember and Ember ), psychocultural (Russell ), culture patterning (Sipes), and t he i nvolvement of women i n war ( Namrnour ) . 201 Quincy Wri ght was the first to engage in quantitative and comparative war fare studies . His Study of War (1 942 ; second e ition 1965) is a massive compendium of theories , discussion , historical detail , a nd descriptive statistics about war throughout human time and space . The bul k of thi s work is devoted to historic warfare , but his attention to primitive war , a lthough brief ar..d dated , i s serious. He has c oded 650 primi tive soci eties for the foll owi ng va riables : regi on ; character of war ( defensive , soci al, e conomic , or political ) ; cli mate (cold , temperat e , hot annual mean temper- a ture s) ; habitat (forest , mountain , seashore , desert , grass- l and ) ; climatic energy (low, medium , high- - based on vari- a t ions i n temperature and humi dity); race (Pygmy , Australoid , ~egroid , Hamitoid , Red , Yellow, Brown , White ); subrace (20 in all ) ; culture (hunters, pastoral s , agriculturalists ); subcul- ture (lower hunters , higher hunters , dependent hunters , etc ., as used by Hobhouse , Wheeler , and Ginsburg) ; poli tical organ- ization (clan , village , tribe, state ); so cia l organization ( sex and age , professional, caste) ; intercul tural relat i ons (i s olated , moderate contact, close contact- -the contacting culture is a civili zation) (Wright 1965 : 544- 51). Wright ' s determination of 11 warlikeness " rests upon t he reasons for t he warfare of a particular group , rea s ons ranked i n a hierarchy of rising expectations, ambition , and i nclinat i ons to engage in war . Thus , war for defense i s a t the lowest l evel , tnen war for revenge , sport , and prest i ge , superceded by war for 202 economic gai n- - s l aves , livestock , land , cu l minat i ng i r_ v:ar to mainta i n or extend the power of a rul ing class . It seems appropriate to r egard people employing this "type of war as the most warli ke of a ll, not only because of their peculiarly favorable attitude toward war but a lso because they re cei ve and i nflict the greatest losses of population fro m war of any primitive people . The high morale which a rmies developed by people of this t ype cust omarily displ ay enab l es them to endure more mutual s l aughter than can the le ss- di s ciplined warrio rs involved i n other types of primitive warfare . Furthermore , the tict ics and weapons used by people of this class are more eff i cient fo r purpo ses of s l aughter (Wright 1965 : 56 1 ) . Wright t hen pro ceeds to tabul ate the i ncidence of each t ype of war with a ll the other vari abl es . I have i ncluded one of hi s t ables , very sli ghtly modi fied , which is still usab le i n t he ant hropology of the current decade (Table 3). Usi ng the 1942 edition of Wright , Broch and Galtung ( 1966) pe r form "tr i vari ate analysis ," c::.~. oss- tabul a ting a ll poss i ble pairs of Wright ' s variable s of world region , cli- mate , habita t , race , culture , subculture , politica l a nd social organization, and i ntercultural r el at i ons wi t ~ the dependent vari abl e "belli gerence "--the presence of war f or economic or polit ica l reas ons . They have found that the whi te state i n close co ntact with other societies i s 100 per cent belligerent ; t hat there i s r i s ing belligerence as one i ncreases in politica l compl exity from clan to village to tribe to state ; that pastoral cultures are highly belli ge rent ; t hat grassland i n North America "dispo ses particular ly to bellige rence ''; t hat South AmericRn states are low i n 203 TABLE 3 . RELATION BETWEEN CONTI NENTS AND WARLIKENESS* Defen- Soci a l Economic Polit- Mean Continent sive ical Average War War War War Total Warlike- 1 • 2 . 3. 4 . ness** No . of Primitive Peoples i n Each Continent Practicing Each Type of Warfare (Percentage of Primitive Peoples in Each Continent Practicing Each Type of Warfare) As ia and 18 ( 12 ) 91 (59) 42 ( 27) 4 (2) 1 55 2. 21 I ndonesia 1 J. orth 7 ( 6) 67 (55) 44 (36 ) 4 ( 3) 122 2. 37 America South 4 (6) 48 (69) 17 (25) 0 (0) 69 2. 19 America ;._f rica 1 ( 1 ) 28 (23) 67 (54) 28 (22) 124 2 . 99 .Australia 0 (0) 75 ( 10 0) 0 (0) 0 (0) 75 2 . 00 Oceania 0 (0) 37 (82 ) 4 ( 9) 4 (9) 45 2. 27 Total . .. 0 ( 5) 346 (59) 174 (29) 40 (7) 590 2.38 * Ada pted from Wright 1965: 551 . ** This figure was obtained by mult iplying each figure by the numbe r at the head of the column , a dding the products i n the row , and dividing this sum by the total at the e nd of t he row . 204 be lligerence (1966 : 35- 36) . The authors have designed an "index of pri mitivity " (o r civilizat i on ) based upon the a ssignme nt of so many points for type of culture , subculture , polit ica l organization , and social organizat i on that , when added , will g i ve a score to a part icul ar society ranging from a low of O (hunting "in it s l oKest form," clans , and age- s ex stratifi cat ion) to a high of 6 (the highest form of agricul- ture wi th stat e organization of spe cia lis t s or cas te s ) ( 1966 : 37) . Cross- tabulating t he i ndex with types of war , Broch and Galtung have gotten "the ir.i.pression that be lligerence is an concomitant of i ncreasi ng civilization" ( 1966 : 37 ; ita lics i n the original) . They posit two- way causation : "increased c i vilization l eads to increased belligerence , which i n turn l eads to increased c i vilization in others ( be cause of ' sur- vival of the fittest ,' homology , diffus i on and simply eradi- cat io n of the more primitive) " (1966 : 41) . In conclus ion , they hypothes i ze that s ince i ncreas i.ng s i milarity between s ocieties i ncreases belligerence due to di re c t competition , such so cieties will f i ght unless some f orm of integration between them occurs ( 1966 : 42) . I n their study , Broch and Galtung deal only with Wright ' s crude variable s and i n per- centage statements of t he cross-tabulations . They use no i nferentia l stat i st ic s . Naroil has 2 studies in pri nt t o date , dealing quanti- tatively with warfare and using inferential statistics , i n which he tests the deterrence hypothesis versus t he arms ra ce 205 hypothesis . I t i s held by advocates of t he former that i ncreasing weaponry de creases t he poss i bility of war be caus e war be comes too costly , whi l e advo cates of the latter argue that i ncreas i ng weaponry i ncrease s the l ikeliho od of war- - rivalr y is i nt erpre ted as intent to wage war . In a n a rms race each side strives without limi t to mili- tary superiority . Ne i ther can be satisfi ed with s i mpl e pari ty , because an underestimated r i val may at any time achi eve a te chnologica l breakthrough that will g i ve superiority .... "We are simpl y prote c t i ng ourselves i n case of a tta ck , but t hey are arming f or war " (Naroll 1966 : 14) . The advocates of deterrence point out that t here is no defense against a nucl ear war , and consequently our best hope is for a stable , credible deterrent . For a det e r - rent to be stable i t must be i nvl unerable so that increased effort agai nst it would be f uti l e . Mutual i nvulnerability would , from this point of vi ew , stabilize the arms race i n a balance of t e rror . This balance woul d then , presumably , allow conflicts t o shift to limi ted wars , and to an eventual s tabiliza tion and finally t o reduction of war (Naro l l 1966 : 15 ) . Tne first study used·a sampl e of 48 (unidentified ) primi tive soci. e t.i es . 2 1 Nar ol l ' s variables are the frequency of war (fr equent or infrequent) and the presence or absenc e of sev- era l traits se en as indicators of an orientation toward war . These traits i nclude ( 1) fir e- and- movement t a ctics : firing at the enemy from a distance and then moving close in fo r hand-to- hand combat , combi ned tacti9s that requi re more c oordination than 1 used a lone and tha t produce more casual - tie s ; (2) flexible surprise tact ic s--re cognizing that sur- prise i s the opt i mum but not the only tactic , tha t if an i mpendi ng surprise attack is di s covered , to persis t i n more 206 formal battle array ; (3) multiple expectations : the order of rising mili tary expectations Naroll has found to be revenge and defe~se , plunder , prestige , and political contro l , a nd he concludes that the more expectat ions a society has , t he more like ly it will be to go to war ; (4) many potent ial enemie s : "The l arger the number of potential foes , the rr.ore likely it is t hat there will be trouble with at least one of them" ( 1966 : 18) ; (5) military readiness : measured by the presence of reconnaissance missions , posted sentinels , or mobilizat i on points ; (6) fortific a tions : forts , fe n ces , man traps ; (7) Western technology : specifically the presence of guns ; (8) repressed hostility : the absence of malicious gossip , quar- reling , or public r i dicule (Naroll 1966 :1 7- 18) . Naroll does not give cell fr equencies or stat istical statements of the find i ngs of his intercorrelat i ons , only identifications as 11 s trong positive , 11 "moderate positive, 11 or "no relationship . " He concludes that his study gives no support to the deter- rence hypothesis , that the only significant correlat ions are between multiple expectations , mi litary readiness , fortifi - cations , repressed hostility, and frequent wars . His strongest relationship is between frequent war and military expe ctations . I n summary , then , ... societies that expect more ki nds of satisfac t ions from successful warfare tend to be societies whi ch f i ght more frequently and which make more preparation for war . 'l1hey a l so tend to be soc i eties wit h large numbers of notential enemi es , tha t use fire-and- movement tactics ~1 966 : 19 ) . 207 Yet Naroll is not clear i n his own mind whether the presence of these vari ables i s the cause or the result of fr equent warfare (1966 : 19) . I n any case, Naroll finds that war orien- tation , specifically the possessi on of guns , has a strong positive corr elation with territorial growth . T1rus societies whose territories increase are character- ized by Western technology and, to a lesser extent , mili- tary readiness . The most plausible explanation is that military preparedness tends to make for territorial expansion . The correlations a l so show a strong tie between territo rial i nstability and mi l itary expecta- tions . So ciet i es whi ch hope for a great deal from war- fare are societ i es whose boundari es a re like ly to change , one way or another . I t is tempt i ng to explain t his as a three-link chain of i nfluence , with war frequency leading to i ncreased expectations , and increased expectations leading to i nstability by making tri bal l and the stakes of warfare . The resul ts here are equivocal , but they do demonstrate tha t warfare is an agent of cultural se lec- tivity , and that the notion that we can best preserve our way of life by throwi ng away our arms i s dangerously naive ( 19 66 : 20 ) . That temptat i on must also l ead us to conclude that t _he way of life one is apparently mai ntaining is a militaristic one . I t is a lso dangerously nai ve to assume t hat one ' s "way of life " i s somehow a separate ent i ty from one ' s military a ctivit ies . Naroll ' s second test of the deterrence hypothesis i s published in more detail , as a pi l ot study "i ntended to try out the cross-cultural survey method of anthropo logy on a comparative s t udy of history" (1969 : 150) . 22 He t ests 30 hypotheses based upon 32 vari ables on a sampl e of 20 soci- eties ( or states) spaced through tin1e i n the randoml y chos en d8cade of 76 to 85 , although the specific century i n whi ch the decade i s pinpointed depended upon availability of data . 208 All vari a bles have been coded for a t arget date , a target sta e wi ~hi n that t i me perio d , and its spe cific re l atio~ship with a target rival . Fre quency of war , territorial gain , and terr i tori al i nstability are dependent vari ables , operationa l- i zed as z scores . Sample members are : Europe State Rival 1 . 225 B.C. Rome Ca rthage 2 . 25 B. C. Rome Parthia 3 . 176 A. D. Rome Marcomanni - Quadi 4 . 376 Ro me Visigoths 5 . 576 Byzantines Persia 6 . 1276 France Engl and 7 . 1376 England France 8 . 1 576 Spain Netherl ands a ./ . 16?6 France Netherlands 10 . 1776 Engl and France Svitzerland 11 . 1 376 Swi ss Confederacy Ki burg 12 . 1476 Swiss Confederacy Burgundy 1 3 . 1 576 Swi ss Co nfedera cy Swi ss Catholi c Russi a 14 . 1476 Mus covy Novgorod Saracens 1 5 . 776 Abbas i ds Byzant i nes China 16 . 125 B. C. Former Han Dynasty Huns 17 . 25 B. C. Former Han Dynas ty Huns 18 . 776 A. D. T ' ang Dynasty Tibetans 19 . 1076 Sung Dynasty Tanguts 20 . 1 376 Ming Dynasty Yunnanese Mongols (from Naro l l 1969 : 154) . 209 Although Naroll argues t hat he has cont r olled for Galton 1 s probl em i n thi s sampl e by using the linked- pair method for both time and space , I cannot help but be alarmed by the i nclus i on of 14 of the 20 cases a s European , despite t he i r spread over 2000 years . The cultural continuity of European hi story i s real, not an artifact , and thus I do not see how one can reas onably control fo r history or diffus ion . Nar oll feels otherwi s e : No tende n cy was evi nced for s uc cessive periods of t ime i n t he same cul tural tradi tioL to re sembl e each other i n the frequency of warfare i nvo lving the conspicuous state ; nor was there any tendency for contemporaneous neighboring cultural traditions to resembl e ea ch other i n thi s respe ct (1 969 :1 58- 59 ). No tendency was evinced for successi ve perio ds of t i me i n the same cultural tradition to resemble each other in t erritorial gain , nor was there a ny tenden cy for cont em- porane ous neighboring cultural tradi t ions t o resemble each ot he r i n this respe ct ( 1969 :1 60 ) . I t i s ent irely possible that t here i s a basic disagreement between Naroll and me as t o what const itute s cultural like- ness and di ss i mila r i ty and thus s olutions to Gal ton 1 s problem . I n any event , bas ed upon coefficients of correlat ion produ ct moment ( quant ita tive variables , i . e . , I s cores) , point biserial (I s core s x qualitat i ve variables) , and phi (qualitat i ve variables) , Naroll found no suppor t for the deterrence hypothes i s , or f or the not ion tha t a defe nsi ve stan ce-- implicitly contras te d with an aggressi ve stance--is less l ikel y to get a state i nvo l ved i n war ( rpb = -. 03) . Naroll i nte r prets the l att er : "I t follows that peace- lovi ng 210 nat i ons are no less like l y to be i nvolved i n war than warlike nat i ons . Hence , or.e must concluQe tha it takes onl y one nat ion to make a war , not two nations , a conclu ion offeri ng no com_ ort whatever to advocates of unilat eral di sarmament " ( 1969 : 1 52) . Yet , s i nce arms make war ri1ore l ike l y , bilateral di sarmament doe s "receive some mode s t err:pirical support " (1 969 : 153) . One would hope s o ; if ne i ther side has arma- ments , war i s a common-sense i mpossibili t y . Naroll cont i nue s to be ambiva lent i n inte r preting the dire ction of what he t a l ks about as t he causal relationship between war fre quency and mili tary preparations, arguing that the l atter causes the former and supporting t hi s with exampl es tha t appear to illust rate that the former causes the l atter (1 969 : 153) . Some of Naroll ' s other pos i tive f i ndings are : the more a ctive are di plomats , the more frequent is war (rpb - . 29) ; ultimatums do not deter war (rpb = - . 21 ). Naroll i nterpre t s diplomacy and ultimatums as "presumably " symptoms "of serious trouble " (1 969 : 158 ) . An inte r esting cluster of posit i ve correlat ions i ncludes fi ndings that rulers over 45 years old are more like l y to be i nvolved i n war ·than younger rulers (rpb = - . 29) , hereditary rulers more likely t han e l ective or self- appo i nted one s ( rpb = . 30 ), and states wi th greater polit ica l centralization more likely than those wi t h l ess centralizat ion (rp,o = . 34 ) . Naroll explaiLs thi s phenomenon as t hat of the rule r who i s more consci ous of hi s status , as contra s ted with h i s role , and i s 21 1 thus more likely to become i nvo l ved in war (1969 : 158) . Yet rulers with more than 9 years of experience tend to gain or lose less erritory than those with less experience (rpb = . 62) . Thus, we have a statistical portrait of a middle-aged hereditary ruler wit~ a great deal of political power going to war often , but for s mall stakes if he is an experienced prince , perhaps very concerned wi th hi s image as a vigorous ruler . Consist ent wi th the 1966 study , Naroll finds a posi- tive correlat ion between territorial growth and the quality of military preparedness and technology (rpb = . 39 ), but t hat civil war within the state and act i ve trade between the state a nd its rival are associateQ with territorial loss sustained by the state ( rpb = - . 44 ; r b = -. 33) . Ambivalence enters .P once again , however , for while trade i s associated with loss , cultural exchanges (undefined here) are associated with t er- ritorial gain ( rpb = . 31) . There are a few other correlations that Naroll finds s ignificant that I shall not present here . I simply wish to make the comment that Naroll seems to be under conside rable pressure to interpret multivariate relationships despite the limitat ions of bivariate analysis using a sample of only 20 cases . The problems inherent in this mi niscule sample make those of Ot terbe in , to be discussed next, look like nothi ng at all . Out of several hundred intercorrelations , Naroll has 212 made a haI'-.dful of statis t ical statements t hat appear to be applicable largely to a handf ul of European states . Otterbein , as co-author and a l one , has published the most material in cross- cultural warfare studies . He works with samples of 50 societ i es , selected ac cording to various randomizing proce dures but i n the long run a ccording to t he avai l ability of data . Otterbe i n and Otterbein (1 965) test ed hypotheses to explain the presence or absence of feuding , i . e ., the occurrence of blood revenge fo llowing a homicide , a nd they c oncluded that feuding occurs when fraternal inter- est groups , i ndicated by the presence of polygyny (whi ch would establish a re sidence group of half- brothers ) ~nd patri locality , are present , but tha t i t is "suppressed " i n societies with a high level of polit i cal integrat i on that engage i n warfare c ont i nually . The vari ables , t he ir sub- classificat ions , and cell freque ncie s for the 50 societies look like this : 213 Level of Polit ical Integration Low High Feuding Feu_di ng Present Absent Pre sent Absent War--Occasional or Never Other 1 5 6 1 1 2 Patrilocal 1* 4* 5 4* 0 4 War--Continual Other 2 6 8 3 6 9 Patrilocal 8* 0 8 2* 6* 8 1 2 1 5 27 10 1 3 23 Otterbe in and Otterbein 1965 :1 479 . This is a small sample ; the cell frequencies , especially when so finely cla ssified , are very small . The addition of 1, 2 , or 3 cases to any class is going to make a difference mathe- mauically. For instance, out of 25 patriloca l societies , 15 have frequent or i nfrequent feuding ; that i s 60 percent . Out of the other 25 societies , that do not have patrilocality, 18 , or 72 per cent , have no feuding . The addition of 3 cases raises a descriptive figure 12 percentage points . The reclassification of any 1 case i n a sample of 50 societies dichotomize d in a 2 x 2 table means a shi ft of 2 percentage poi. nv+ s . Thi s is a great dea l of potential error . The geo- graphical re presentativeness of t he sample i s Afr ica, 8 societies , Circum- Mediterranean 2 , Eas t Eurasia 7, I nsular Pacifi.c 11 , North America 12 , and South Aruerica 10 . Of t hese s ociet ie s , the Bemba and the Yao belong to the same 214 World Sampling Province , as do the Murngi n and. Ti wi , and t he Abipon and Mataco (Murdock 1968 ) . All but t~e Tiwi are co ded as havi ng continual warfare and thus constitute 5 of the total of 28 societies so coded . I t woul d have been useful if Otterbe i n and Otterbe i n had enumerated the specific societies i n the crucial 6 positions marked by an asterisk i n the table above , to a llow further checki ng . Otterbein (1 968 ) continued his quant i tative research , extending his i nterest from relationships between feudi ng and general warfare to concentrate on the nature of i nternal war , that is , uarfar e between culturally similar but politica lly autonomous communities . I t is demonstrated that fraternal i nte re st groups and unauthorized raiding parties infl uence t he frequency of i nternal war in uncent ralized political systems , but not i n centrali zed ones . I t i s a l s o shown that the frequency of external war (warfare between culturally different political communit ie s ) does not influence the frequency of i nternal war (Otterbe i n 1968 : 277). That is, where there are groups of related males , t hese males will act i n concert to take revenge for a homicide ( feud ing ) and to raid other poli t ica l communities as small- s ca l e mili - tary organi zations . I ndivi dua l warriors will ge t a r aiding party together . As f or the organizing principle for such groups , Otterbein found correlations between patriloca lity, the obvious principle , and continual or freque nt i nternal war to be i nsi gnificant wi th a very weak phi correlat ion coeffici ent of . 14 . Surprisingly , polygyny has a higher cor- relati on wi .h internal war , with a modest phi of . 31 : 215 I nte r nal War Continual or Frequent Infre quent Polygyny Absent 1 3 1 1 -24 Present 1 5 3 18 28 14 42 x2 = 3 . 93 p < 0 . 05 (Otterbein 1968 : 281) . Yet fraternal j_nterest groups as indi cated by po l ygyny are no better a pre dictor of the frequency of internal war than the va::-iable I nitiating Party dichotomized into "Anyone " and "Offi cia l . " For uncentralized political systems , either fra- ternal interest groups or anyone as initiator may be associ- ated with internal war , but in any case , centralized pol it i- cal systems with officials i nitiat i ng war have more frequent i nternal war . In centralized political systems , "offici als " rather than "anyone " init i ate warfare . The i nt ercorrelation is high ~ but Otterbein interprets the i ntercorrelation as sup- porting the hypothesis nthat the higher the l evel of polit i - cal complexity , the l ess like l ihood of war being initiated by anyone in the political community . . . . Apparently central- ized polit ica l systems are able to prevent unauthori zed part i es , which would i nclude frate rna l inteiest groups , from engaging i n war " (Ot t erbei n 1968 : 282 ) . But it turns out that 216 internal warfare i s more , not le ss , frequent i n centrali zed poli t i es . The only conclus ion Otterbei n draws from this is tha t grou ps engaged i n i nternal war mus t be authorize d i n centrali zed polities . This appa rent i ncrease i n i nternal war with central ization seems to go against the ca se Otterbe i n seems to be argui ng--that centra lizat ion beings i nternal peace by controlling inherent l y violent male asso cia tions-- yet he persists wi th a non sequitur : I have ... shown that uncentrali zed political systems in and of t hemse lves are not prone to more i nternal war than are centralized polit ica l systems ; in fact , they are prone to l ess . Thus it seems t hat i t i s t he existence of unauthorized r a iding parties , r ather than the absence of a cent r a li zed politica l system , that l eads to i nternal war (Otterbe i n 1968 : 283 ) . Since he means by 11 unauthorized ," "anyone " or nonofficials , and uncentralized polit ical systems by defi nition mean a ';} sence of' officials , the hypothes i s appears tautologous . Anyhow , the test of the hypothes i s "that societ ies i n which a nyone can i ni t i a te war are more likely to have internal war t han societ i es i n whi ch an official i nitiat es war " (1 968 : 283 ) produce d i ndicat ions to the contrary : ... i t appears that centrali zed polit ica l systems in which an official is t he only one who can i ni t i at e war are more likely to engage i n i nternal war than are centralized politi cal systems i n which anyone can i nitiate war. The data seemi ngl y provide support f or the i nterpretat ion that in uncentralized politica l systems wars are i nit i- ated by unauthorized parties and officials are the par- ties who try to mai ntain peace , while i n centralized po l it i cal systems wars are i nitia t ed by officials (Otterbein 1968 : 284 ) . 217 To exp l ain further intercorrelat ions that apfear to i ndica te that centralize d politica l systems wi thout f rat ernal i nterest groups , i ndicated by the absence of polygyny , are "more prone to i nternal war , 11 Ot terbe i n a sserts tha t A possible interpretat i on i s that s i nce societies without fra ternal i nterest groups l a ck conflict-- specif ically feuding--within the politica l community, their officials are more willing and able to engage in i nternal war ( 1968 : 284 ). There is certai nly no evidence in the i ntercorrelat i ons t o war rant such an interpretation , other t han transformation of the Sumneri an hypothesis of an inverse re l at ionship between i nternal and external viole nce . I ndeed , Otte rbe i n ( 1968 : 285 ) goes on to test the amity- enmi ty complex , tha t the more frequently political com- muni t ies fight those who are culturally di ssimilar , the less like l y they are to fight political communities that are cul- turally s imilar to themse lves . As s caled , not only are the correlat ions not significant , t hey cannot support interpreta- tion as tendenci es , e . g ., phi coefficients of . 05 , . 11, . 13 , . 14, .17, and . 29 . Nor does supplying an i ntervening vari- able of centralization tease out any s i gnificant i nverse relationship between i nternal and exter na l war . One inter- esti ng hypothes i s . that does have statistica l support i s given : The assumpt ion made by many writers that revenge and retali ation pl ay a prominent r ole i n war can be submitted to empirical test if it i s f ormul ated as follows : the more frequent ly the politica l communitie s of a cultural 218 unit are attacked , the more frequently they will attack other societie s (Otterbe i n 1968 : 285) . Distribution of cases looks like this : Fre4uency of Attacki ng Infrequent Cont i nual or Frequent Frequency of being attacked Continual/Frequent 5 18 23 I nfrequent 1 5 9 24 20 27 . 47 cp = 0.41 x2 = 7 . 98 p < 0 . 0 1 (Otterbein 1968 : 286 ) . It is not clear , however , which is the dependent and which i s tne i ndependent vari able . Does being attacked lead to attacking or vice versa or both? If one were to accept the statistical validi t y of this study , one could perhaps as credibly interpret the results as showi ng that the fewer the number of people authorized to initiate war, the more like l y that both i nternal a nd external war will occur , and that violence escalate s into more violence . Furthermore , Otterbe i n ' s conc l usions are confusing . Reiterating the findings of the 1965 study about frate rnal i nterest groups , he says (1968 : 287 ) , "I n both studies it was found that the level of political complexity had no signifi- cant influence upon either feuding or internal war . That is , centralized political systems are just as l ikely to be 219 cl:.aracterized by feuding and i nternal war as uncentralized political s ystems . " This is patent l y a contradiction of what he says about his findings in 1965 : Not only i s there a strong re l ationship between war and the absence of feuding in hi gh level societies [ cent 1·al - ized polit i es] , but in low l evel societies [uncent ralized polities ] war and feuding are positive l y correlated . These re sults i ndicate t hat a s ocie t y which has a high level of politica l i ntegrat ion is i ndeed capab l e of pre- vent i ng the i nt ernal conflict which woul d be de t r i mental to its welfare . Although it would seem judicious for any society engaging i n war with i ts ne i ghbors to have i nter- nal cohesion , soci et ie s with only a low l evel of politi- cal authority apparently cannot control the feuding whi ch is engendered by the presence of fraternal intere st groups ; in these socie t i es war and feuding go hand- i n- hand (Otterbein and Otterbein 1965 : 1478) . The rest of Otterbein ' s 1968 conclusions are r i ddled wi th c ontradictions : To summari ze , in uncentralized politica l systems frater- nal interest groups are a determinant of both feuding a nd internal war , whereas i n centralized polit ical systems fraternal interest groups are a determinant of feud i ng but not internal war . I t has been argued in both studies that offi cials i n uncentralized poli t ica l systems are unable to prevent fra ternal i nterest groups from engaging in e i ther feuding or i nternal war ; on the other hand, i t i s diff icult to unde r stand why of ficial s i n cent ralize d polit ica l systems- - who apparently can prevent unaut hor- i zed raiding part ies , i nc l uding fraternal i nteres t groups , from engaging i n i nternal war- - would permit fra- ternal i ntere s t groups to engage i n feud i ng (Ot terbe i n 1968 : 287) . Contrast this interpretation of his findings with t he fo llow- ing , found i n the paragra ph i mmediat el y succeeding that quoted above and after sta tement s tha t there i s no relat ion- s hip between war and feuding , or i nternal and ext ernal war . However , when the rela tionship between war and feuding was controlled for level of political compl exit y , a strong r e lationship between war a nd the absence of 220 feudi r g was found i n centralized political systems , but i n uncentrali zed poli t ica l ystems war and feuding were positively correlated . Apparent l y officials i n central- i zed pol itical systems intervene to prevent development of feuding only when the soc i ety i s engaged i n war . On the other hand , controlli ng for l eve l of poli ti cal com- plexity does not affect the relationship between external and i nternal war . I n other words , both uncentralized and centralized political systems within a cultural unit engage i n int ernal war wi th the same frequency as they do external war . Seemingl y off icia l s i n cent rali ze d poli ti- cal systems do not unite , and t hereby eliminate i nternal war , when engaged i n external war (Otterbein 1968 : 287- 88 ) . Otterbein includes the l ast i nterpretation i n h i s conclus ions af t er te sting i t as a hypothesis and f i nding no support , even tendencies, in his coefficients of associat i on (1 968 : 286- 87) . I n the end , Ott erbe i n falls back on case illus tra tion , non quantification , as evi dence for his hypotheses , specifi- ca l l y the Yoruba wars of the 19th century . I n my est i mat ion , t he only way his hypothesis about the amity- enmity compl ex c ould be demonstrated would be to deal only wi th ca s es of polit i cal centralizat ion where p ol it i ca l unit and cultural unit are i dentica l . Then , by defini tion , any war that would erupt internally woul d be revolution or civil war . By a deft stroke , increasing internal peace a nd i ncreasi ng centraliza - tion would coincide and thus remain consistent with mid- t went i e th century demo cra tic poli t ical philosophy tha t t he state i s the pana cea for conflic t , and that anarchy (in its pure sens e) allows the natural i nclinat i on of male groups to fight to flourish . Thi s judgment may be harsh , but Ot terbein refu.ses to admit the seriousness of non- significance i n his 221 correlations , which he must do to play the numbers game fairly , and persists in representing his hypotheses as proven, although i ndirectly and wi th a good deal of now- you- see - it - now- you- don ' t i n his writing . With regard to the mechani cs of this study , the sample i s again 50 cases , distributed geographically: Africa , 10 s ocietie s , Circum- Mediterranean 4 , East Eurasia 8 , I nsular Pacific 9 , North America 10 , South America 9 . This time , none of the 50 seem to be cultural duplicates . Because of uneven data , most of the N' s for specific inte r correlations are lower tha n 50; t hey are usually 36 or 42 and , once, 47 . With de creasing sample s i ze , the mathematical effect of a single case increases , and with finer scaling , the number of cases in each cell diminishes and may even di sappear . Inspect the following example. 222 Uncentralized Centralized Political Systems Polit ical Sy~tems Int ernal War Internal War Cont i nual Co ntinual or Frequent I nfrequent or Frequent I nfrequent I nitiating party : Official 3 6 9 9 2 1 1 Anyone 1 2 2 14 1 2 Total s 1 5 8 23 10 3 13 Grand Total 36 cp = . 54 cp = -. 27 x2 = 6 . 63 x2 = . 97 p . 01 n . s . (Otterbein 1968 : 283) . For these intercorrelat i ons , Otterbein is worki ng wi th sample sizes of 23 and 13 . We are certai nly faced with the dramatic effect of a single case . Otterbe i n ' s research culminat es in a study t hat tests relationships between political centralization and military activ i ty : All the above theories [ about the evolut ion of the stat e and war] have i n common the not i on that societ ie s be come socially and politically more developed through time . As societie s evolve , they come to wage war i n mor e efficient ways . Sometimes war i s seen as producing the evolution of societies ; someti~es it i s the politica l l evel of the soci eties which is seen as be i ng re sponsible f or the type of war waged . I n either i nstance , level of politi ca l centra lization and degree of military efficiency are viewed as being functionally re l ated . Thi s study tests a series of hypotheses which relate leve l of politi cal cen- tralizat ion to various aspe cts of warfare . I n most instances i t will be shown that more efficient mi litary • 223 practices are associated with centralized political sys- tems . Support is also provided for the general hypothe - sis that as societies be come politically more central- ize d , they wage war in a more efficient manner . The study also demonstrates that societies which wage 1-rar e fficient ly are likely to be militarily successful (Otterbein 1970 : 2 ; italic s added) . Military efficiency i s measured by the presence of specific military practices that , in Otterbe i n ' s e st i mat ion , i mprove a f i ght i ng force ' s chances of defeating it s opponent, e . g . , shock weapons over projectile weapons (bombs over bullets) and "a high degree of subordination" over a low degree of subordination i n organi zation of personnel (general s over warrior chiefs) . Seventy-five per cent of the societies in the 1970 sample have a high degree of subordinat i on . Each of the more efficient military practices is associated with cen- tralized political systems, since such systems are assumed to be higher on the evolutionary s cale . The pre sence or absence or the types of these and othe r traits--mi l itary organiza- t i on , initiating party , diplomatic negotiations, tactical systems , protection, field fortifications , cavalry , fortified vill ages , causes of war--are coa lesced into a military sophistication scale , and an efficiency rating i s derived by dividing the total presence s of the 11 military practices considered to be more efficient by the number of absences . Thus, if a society has a professional military organization with a high degree of subordination but no cavalry, it gets 2 points for presences and 1 for absences . Sophisticated societies are those that have a rating of . 50 or higher ; 224 uns ophi st icated one s , . 49 or l ower . Out of a total of 46 societie s , 27 rated a s being militarily unsophisticated and 19 as s ophisticated (Otte rbein 1970 : 70-74 ). The sample of 50 societie s this time is drawn with respect to Murdock ' s 6 geographical regions of the worl d , 1 each of the 60 culture areas . In the long run , admitting such limitat ions as avai l ability of data in the English l an- guage in Lawrence, Kansas , duri ng the spring of 1965 , an ava ila~ility sample of 50 societies was eventually drawn . No representative for 10 culture areas was drawn on any cri- teria . For the final sample, the geographical di s tribution i s : Africa 10 societies , Circum- Mediterranean 4 , East Euras ia 8 , I nsular Pacific 9 , North America 10 , South America 9 . Of the 50 , 4--the Copper Eskimo , Dorobo, Ti kopia , and Toda- - did not have any kind of military organization , due to i so l ation , whether indigenous or refuge (Otterbein 1970 : 12-1 4 , 20 ). Let me set forth Otterbein ' s stri ng of hypotheses and their stat i stica l test results . (1) "The higher the level of political centraliza tion the higher the degree of military s ophisticat ion11 (O tterbein 1970:75) . For the intercorrela- tions , the variable politica l systems are dichotomized i nto centralized and uncentralized ; t he variable military sophis- ticat ion i s dichotomized into low and high . Statistics used are point biseri al with a (1-ta iled? ) t test worked on the efficiency ratings f or every society in each cell, and phi. 225 Military Sophist i cation High Low Political Systems Centralized 1 3 3 1 6 Uncentralized 6 24 30 N = 46 cp = . 59 x2 = 16 . 1 5 p = . 001 (Otterbe in 1970 : 74 ). ( 2 ) 11 The higher the degree of military s ophist icat ion , the higher the casualty rates " [ be cause those weapons that kill the most enemies a l so bring the user closer to the f i ghting , givi ng the enemy greater opportunity to inflict damage on personnel ; the Zulu adoption of the short stabbing sword and the c oncomi tant r ise i n casualties i s an example ] . Casualty rates are dichotomized i nt o high (1/3 or more of the s oci- ety ' s combatants killed ) and low . Mortality Rates High Low Military High 8 7 1 5 Sophis t ication Low 5 1 3 18 N = 33 rpb = . 48 t = 3 . 04 p < • 01 er= . 26 x2 = 2 . 24 n . s . (Otterbein 1970 : 83) . ( 3) 11 The higher the degree of military sophistication, t he more likely that the poli tical communities of a cu l tural unit 226 will engage in frequent or cor.tinual internal war [warfare between polit ical communities of the same cultural unit] ." This hypothesis is justifi ed on the a s sumption that military sophistication must come about through frequent wars, i . e . , multiple opportunities where weapons are tested and retained or di s carded . There is a curious twist here : i n no way are we to conclude that military sophi st ication leads to frequent wars , but i nstead the causal arrow points in the other direc- tion . Military sophist icat ion , as Otterbein presents it, i s the dependent variable . The i ntercorrelat ion based upon 40 cases was found non-s i gnificant , so this hypothesis is dis- carded ( Otterbein 1 970 : 85- 86) . ( 4) "The higher the degree of mili tary sophi stication , the more likely that the polit ical c ommunities . of a cultural unit will engage i n frequent or continual offensive external war " (Otterbein 1970 : 88) . The difference be t ween this hypothesis and (3) i s that external war is warfare between political communities of different cultural units , and here the so ciety in question is the aggressor . This i ntercorrelation was found to be non- significant with a phi of . 17, but the point biserial corre- l ation of . 32 was s i gnificant at p < . 05, with unspecified degrees of freedom . ( 5) A thi rd variation , "The higher the degree of mi litary sophistication , the less like ly that the political communities of a cultural unit will be attacked 11 (Otterbe i n 1970 : 90) , turned out to be nonsignificant on all counts , thus providing non- confirmatj.on of the deterrence 227 hypothesis , favorite of most col d war politicians , mi li tary personnel , and muni t i ons manufacturers . Thus the only even f a i nt l y posi t i ve corre l ation i s that societ i es wi th efficient means for killing peopl e are likely to be the attackers and i n the subsequent war l ose a great many f i ght i ng men . The fina l variab l e of concern to Otte rbei n is milita ry s uccess- - what i s i t and do e s military sophist ication bring it about ? As a standard measure of success that i s not based di rectly upon casualty rates , su ccess i n i ndi vidual batt le s , or mot i ves , Otterbe i n uses territorial boundaries of a society : do they expand , rema i n unchanged , alter but retai n the same area , or cont ract? Military succe ss i s measured by territorial expans ion . It i s di ff icult to i magi ne tha t a politica l community which i s be i ng defea t ed by it s neighbors could be expand- i ng territoria lly , unless it is be i ng dri ven i nt o mar- gi nal , uni nhabited l and (Vayda 196 1) . The four s ocieties i n this s tudy which do not have mi litary organi zations- - Copper Eski mo, Dorobo, Tikopia , and To da--were appa rently driven f r om more desirabl e l ands (Ot te rbe in 1970 : 93 ) . Of 39 s ocie ties in the sample , 13 are expandi ng terr itor i- ally, 24 have remained unchanged , 3 a re shi ft i ng , and 3 are c ontracting . Thus , ( 6 ) is genera ted : "The hi gher the degree of military s ophist i cation , the more like l y that the politi- ca l communit i es of a cultural unit will be militarily suc- cessful " (Ott erbe i n 1970 : 94 ) . By l umping constant and shift i ng boundaries with cont racting ones , the intercorr e- l ation t able l ooks like thi s . 228 Mili tary Success Terri tory Terri t ory Not Expanding Expandi ng Mi litary Hi gh 10 9 1 9 Sophi st i cation Low 3 1 7 20 N = 39 cp = . 40 x2 = 6 . 21 . 01 < p < . 02 rpb = . 44 t = 2 . 96 p < . 01 (O t t erbe i n 1970 : 95 ). Apparent l y te st i ng the traditional assumpt ion tha t polit ica l central ization and military success g o hand i n hand ( bands and tri bes by defi nition do not have "true war ," but chi ef doms and s t ates do , t here f ore pri mi t ive s ocieties cannot do as well as s tates ) , Ot terbei n f ormul ates (7): "The high~ r the leve l of poli t i cal centralizat ion, t he more like l y that the poli ti cal communities of a cul tural uni t will be mili- tari l y successful" (1 970 : 97) . The resul ts were qui te a sur- pri se : the phi value was onl y .1 2 . Fur ther s ubclassi f ica - tion of the vari ab l es (1970:1 00-1 01) fa i led t o a lte r the f i ndings that only the efficiency of a group ' s t e chni ques for kill i ng i s associated with expandi ng terri tori a l dominat i on . Thus in the end i t i s armaments and tactics that i nfluence the outcome . While i n the test of (1) above , there i s a s t rong correlation between mili tary sophi st ica tion and polit - i cal centralizat i on , and between (6 ) mili tary sophi st icat i on and military success , there is no correlat i on between po l it- i cal centralizat i on and military success . That i s , while 229 Otterbein may i dent ify centralized political organization as a more evolved form than uncentralized, and efficient methods of killing as more highly evo l ved than i nefficient ones , they do not necessarily evolve together . If it is true that a political community will probabl y defeat any political community which wages war in a manner less s ophisticated than i ts own , then it i s t o be expected that some uncentrali zed political communities wi ll be abl e to defeat some centralized political communities. Al though this study , because of the method used i n draw- i ng the sample does not provi de direct evi dence that there are uncentral ized politi cal communities wi th hi gh mi l i tary sophistication s cal e scores which have defeated centralized political communit i es with low military scphistication scal e· scores, i t does provide i ndi rect evidence that they may occur . The fact that a cultural unit is composed of centralized political communities does not ensure it of military success (Otterbe i n 1970 : 10 7) . Thus the hypotheses offered as proven in the introduction t o his study ( 1970 : 2) are not, in fact , linked in the way they appear to be . I t looks to me as though mi litary sophist ica- tion may be the independent variabl e . I s i t possi bl e that the relatively small nuJnber of people who make de cisions i n cent rali zed polit i cal systems lead their people i nto frequent or continual wars , l ose great numbers .of them , and l ose i n the long 1un? Can we tentat ively e xplain the fall of states , the eventual demise of every empire that has existed , so often falling before the "Barbarians who ride out of the Nort h "? Military sophistication also may respond to a "Law of History '': Increasing sophistication does not deter war , iv generates it ; but wi thout eff icient force of arms , a 230 society cannot expand and thus i s faced with mi litary fail- ure , since Ot terbe i n uses only expansion as an indicator of success . Otterbein does not offer these i nterpretat i ons or i ndicate pursui ng t hem , possibly be cause they go 1:,gai nst tra- di tional politica l and evol ut i onary t he ory , which sees whole complexe s of social i nst itut i ons sti cking together , and whi ch places a high value on s ocial spe cia liza tion and po l i t ical centralizat ion , without exploring the possibi lity that the evolution of higher biological forms i s predicat ed on the existence of variation, plast icity , and non- specialization . That i s , cultural speci a l izat i on and concomitant centrali za- t i on are to human beings what mons trous ant l ers were to the I r i sh El k . Thi s may be the human paradox , generated by dearly held beliefs . I n conclusion , then , Ot terbein has found positive stat i st ica l correlat ions between efficiency i n killing the enemy and territo r i al expansion . Adhering to these variables are freQuent war and high mo1t ality of combatants . That i s it . That i s the nature of any evol ution of war . Otterbein offers his findi ngs wi th a different cast : It can be argued that this study provides evidence for the evolution of war , if t he following assmnpt ion is made : namel y t hat for any pair of alternat ive milita ry practices , the more efficient practice i s more evolved than the le ss efficient pra ctice . The bas i s for making this assumpt i on is that the more effici ent military prac- tice confers a survi val advantage upon a political commu- nity by increasing the like l ihood t hat i ts military organi zation will defeat the military organizations of other politic:a l communities , thus also i ncreas i ng the likelihood that the political communi ty will be the sur- vivor i n intersocietal struggle s (1 970 : 105) . 231 I rr.ay remind the reader that the rhetoric of f i ght or die out in thi s case equates no territorial expansion with nor.- survival . Let me again quote Naroll i ~ his Foreword to Otterbeints study : Of the many observations Otterbe i n makes on the subje ct , the one that most i mpresses me i s hi s assertion that in war the test of fitne ss is applied , not just to milita ry pract ice s , but to societies themselves . The u l timat e test of f itness , of course , is survival . And no matter how well adapted a society may be i n othe r respects , if i .t proves unable to c ope wi th i ts enemies it has failed i n i ts overall adaptation and must give way . He re li es t he real i mpetus to evolution in warfare and i n a ll aspects of culture that are conne cted with war .... Coldly viewed , warfare has enormous e cologica l s i gnifi- cance . I t i s concerned , after all , with a most vi tal aspect of a society ' s environment - - its enemies . I f waged successfully, war means the pre servation of a society 1 s i ntegrity and i ndependence , and the defense or even i ncrease of i ts territory and resources . If waged unsuc- cessfully , warfare may mean defeat , subje ction , or even extermi nat ion . Thi s is ecology in spades ! (Otterbe in 1970 : i v- v ) . It i s also Soci al Darwini sm i n spades! Otterbein performs one other i ntere st i ng exercise wi th his sample : a scale of the "causes of war" or what Naroll ( 1966) calls military expectat ions , i . e ., the reas ons why a society goes to war . While Naroll i dent ifies military expe c- tations as those given by the pe ople themselves and disavows the possibility of tal king meani ngf u l ly ab out causes , Otter- bein equates cause s with "goals ," whi ch may confuse cause with causal effect , and he all ows his coders t o de cide what t he goal s of a society are (1 970 : 64 ) . Otterbei n concl udes that his replication of Naroll ' s 1966 s cale f i nds the same 232 results , and that both refute Wright ' s 1942 scal e , where he determi ned that the orde r of rising expectations i n war was defense , social , economic , and poli t ical . The order of the s ca l es follows , with Otterbein ' s lurnp i Lg of categor ie s to gain comparability wi th Naroll 1 s i Ldi cated . Wright ' s Scale Naroll ' s Sca le Otterbein 1 s Sca le polit ical politica l control subjugation and tribute economic prestige trophie s and honors social pl under l and defense revenge and defense plunder r evenge defense Otterbe i n ' s pos itioni ng of defense and revenge as pr imary did not come out of his coding of the data : ... although I recorded what I t hought were all the i mportant causes of war , only 29 out of 46 s ocieties fought for either revenge or defense . Apparen t l y if I f ound several more sophist icated reasons for war , I did not seek and record ment ions of revenge and defense . Therefore, I assume that all t he societies in my sampl e (e xcept those without military organizations ) fought either for revenge or defense (1 970 : 65 ) . Otterbein ' s justificat ion for this i nference is apparent ly that defense and revenge- and- defense appeared first i n 1 \ lright ' s and Naroll Is scales , an odd practice if one i s sup- posed to lie replicat i ng and therefore t esting j ust t hose s cales . I am guilty of the same practice in my fir s t cross- cultural warfare study , with 35 societies ; I made the 233 assumption that revenge and defense had to be there , even if not mentioned specifically in the data--after all , everyone fights back and wants to get even , right? But not neces- sarily so . The interpretation of Naroil ' s and Otterbein ' s scales , then , i s that either revenge or defense is always present i n war ; where there are reasons of prest i ge , there are also always economic reasons ; where there are goals of political control , there are also soci a l , e conomic , and revenge- or-defense goal s (Otterbein 1970 : 65- 67 ). This concludes my dis cussion of Otterbein ' s contribu- tions to warfare s tudies . I may only point out again that while the 1970 study i s more s ophisticated and di s crimi nat i ng in selection of sample , doubt les s under the i nfluence of lfaroll, we are st ill confronted with the problem of small sample size and the mathemati ca l effect of single cases . May I also point out for the last time that even if Otterbein ' s corre l at ions are statistically valid , they are subje ct to various interpretation, and Otterbein has not poi nted out the i nstances where they seem t o refute orthodox polit ica l and s i mplist ic evolut ionary theory . Ember and Ember (1 97 1) perform a series of tests of explanations as to why matrilocal residence occurs rather than the patrilocal mode , specifically the traditional expla- nation i n anthropology that sexual division of labor i s the pre condition for post- marital residence . They refine Murdock ' s division of l abor variables i n the Ethnographic 234 At l as to these attri butes : women do more than men overall i n the s ub isten c e activ i ties ; neither sex predomi nate s ; men do more t han women . They l ump the residence coding i nto 3 at tributes : females localized , ne ither sex loca lized, males l oca lized . They have found no significant relationships in t h e crucial cells of the 3 x 3 t ab le on a world- wide sample of 288 societies . I n retest i ng Driver and Massey ' s findings for North America \1957) , one of the orig i nal case s upon wh ich d i v i s ion of l abor was found to be the s i gnificant con- dit ion for post - mari tal re s iden ce, Ember and Ember ' s result s d 1 p licate those of Dri ver and Massey but only for North Ameri ca . The r e l at i onshi p between div ision of labor and res- i de ~ce d oes not hold up o n a world-wi de sample , and it vari es wi de l y i n the othe r geographic regions of the world . Th e 2 examples t he Embers g ive are a nonsignifi cant relationship in South America , and a nearly s i gni f icant negative re l ationship i ·ri Oceania ( 1971 : 575) . I n l ooking for other possible determin ers of res i - dence-- s pe cificall y conditions that give one s ex greater social status than the other- - Ember a n d Ember fi nd polygyny , herding , slavery , and multiloca l politica l organization to be significa nt ly assoc i ated wi th male l ocal ization on the large sample , but on l y herdi ng and multilocal politica l organiza - tion are good pre d ictors of male loca l izat ion . The Emb ers use the term "pre dic ors " a lthough the i r statist ic s are on ly Phi and Fisher ' s Exact fo r probability l evels . They do not 235 use a coefficient of predictability . A last variable , war- fare , could not be based on the Ethnographi c Atlas s i nce it does not code that variable , so Ember and Ember use Otterbe i n and Otterbein ' s 1965 sampl e and find no relationship between cont inual warfare and male localization . Ember and Ember conclude that an explanat i on of resi dence t hat is based upon sexual social status i s not very s trong and that , i n effect , t here may be unidentified i ntervening variables ( 1971 : 576-77). Using the Otterbeins ' 1965 sampl e , they proceed to i dentify continual warfare as the intervening variable in t he associ- ation between di vision of l abor and residence : where warfare i s cont i nual but men do mo s t of the subsi stence work, resi- dence is usually patrilocal ; wher e warfare is co nt inual but women do most of the work , r e s i dence i s usually matriloca l ; wnere there is no continual warfare , residence still tends to be patrilocal regardless of which sex does most of the sub- sistence work . The task remaini ng i s to i dent i fy what cir - cumstances i n warfare influence whether or not the men could kee p up with the i r subsistence work . They make an interest- i ng [and erroneous ] assumption that the normal di vi s i on of subsistence labor is patridominant be cause women are taking care of t he children and "other uni versally ascr ibed dut ies around the home " (1971 : 578 ). Thi s assumption gets them into some diffi culty l a t er in their i nterpretat i ons , spe cifi cally trying to explain the cases in which women do as much as men without recent warfare as due to "culture lag " (1 97 1 : 581) . 236 On their own warfare sample of 22 s ocieties (Afr ica 10 , Circum- Mediterranean 1 , East Eurasia 0, I nsul a r Pa cific 3, North America 6 , South America 2 ), Ember and Ember test their fi r st hypothesis that "men will do mo r e than women in subsistence unle ss warfare prevents t hem from doing so " ( 1971 : 579 ) . Circumstances seen as preventive are t hose wnere ... (1 ) warfare occurs fre quently all year r ound ; (2) warfare occurs at l east once every two years but poten- tially can occur at a ny time (as indicated by year- round sentries or fortified villages ); ( 3) warfare occurs fre - quently only at certain times of the year and work i n the dominant subs i s te nce activity has to be done at that time ... ; (4) warfare occurs at leas t once a year a nd is at l east sometimes offens i ve, the travel to and/or the fighting itself i s a t least sometimes long in durat ion (that i s , more than a day) , and in additibn work in the dominant subsistence activity has to be done while fight- ing takes place outside t he communi t y ( if warfare occurs only within the confi nes of the community , it will dis- r upt a l l subsistence l abor and he nce will not prevent the men from do i ng more t han the women) (1 971 : 578 ). Under a ll other conditions of warf are , men will assume their normal dominance in subsistence l abor . Ember ' s and Embe r ' s predict ions of t he divis ion of labor a ccordi ng to the se cri- teria i nterco r relate d with the a ctual di vision of l abor for t he 22 societ ie s are marginally s ignificant (FE= . 074) . They next t est the i nfluence of r esidence upon war- fare , with the inte rveni ng variable assumed to be desce nt, i n the second hypo t hesi s that matriloca l communities will have wars only with other s ocieties ( external warfare ) because t he ma l es i n the matrilocal communi ty will be fighting consan- gui neal kin i n any internal warfare, but patrilocal communi- tj_e s will fight ei ther other societie s or othe r cornmuni ties 237 of their own cultur e, because the consangui neal ties of the men are all i n their home community . This relationship was highly significant--on a sample of 18 ( q, = . 79 ; FE=< . 002) . The geographical representation of this sample is 7 Africa societies , Insular Pacific 2 , North America 5 , South America 4 . Uneasy, Ember and Ember replicated this i ntercorrelation on a larger sample of 33, and again found the relationship to be significant (~ = . 65 ; FE< . 01) . The geographical distri- bution of this sample is Africa 12 , Circum-Mediterranean 1, I nsular Pacifi c 5, North America 9 , South America 6 . At this point Ember and Ember have 2 i ntercorrelations &nd 3 variables : war , division of labor , and post - marita l r es idence . The second and the third variables do not inter- I correlate significantly , or they did not i n the original l a rge sample . Therefore, the authors proceed to i nterpolate these variables through verbal , rather than statistical , reasoning . I quote their exegesis at length because it is i nteresting i n argument and inference and because spme of my own intercorrelations are referable to those of the Embers . The i nitial state is the situation where warfare i s pres- ent . (Judging from our sample data , this is the statis- tically normal s i tuation : most societies as of the time of description either were still fighting or had only recently stopped fighting , the latter being the case usu- ally because of pacification by a colonial or i nternally conquering power .) If the warfare i s at least sometimes internal, there i s only one possible resultant state for residence but two possible resultant states for division of l abor . The one possible state for residence , when warfare is present and at least somet i mes internal , i s patrilocal . Again judging from our data , the fact that warfare is at least sometimes internal appears to require 1 238 patrilineally related males to be localized after t heir marriages . Or, i n other words , i f fighting occurs betwee n ne ighboring communities , familie s woul d want to keep their fighters at home for protection . As for division of labor , the two possible states are produced by the interaction between the int ernal warfare a nd the nature of subsistence work . If subsistence work has to be done while fighting occurs , the normally patridominant division of l abor will be disrupted and females will come to predominate ( or at least contribute equally to subsi s - tence) .... On the other hand , i f work does not have to be done while f i ghting occurs , the males will continue to predominate in the di vision of labor . Turning now to what happens when a pure l y external pat- tern of warfare emerges , there are two possible resultant states for division of labor and two for residence. I f subsistence work does not have to be done while the purely external warfare occurs , the normally patridominant divi sion of labor will be disrupted such that the females will c ome to predominate (o r at least contribute equally to subs istence), and the pattern of residence wi l l change to matrilocal . This l atter change will occur because, in the absence of local f i ghting (internal warfare ), patri - lineally related males do not have to be localized; that is , there would be no need to keep the men at home after marriage for protection . Thus , because the females sup- port the kin group , they rather than the males would be kept at home after marriage (1971 : 584- 85 ). Note the confus i on of cause and causal effe ct and the evi dence given to support it--war is primary because most s ociet i es were experiencing it at time of contact ; note also the disre- gard for geographical variat i on , which made the f i rst part of the research a significant contribution . The absence of this variable and the distribution by inspection of t he societie s makes one wonder if not controlli ng for geography was a tac- tical decision or based upon the assumption that i f a corre - lation appears to hold on a world-wide sample , geographical variat ion is irrelevant . The i ni tial reply, of course , is to 239 quest i on whether or not the world- wide sampl e i s really "world- wide ." I n summary , Ember and Ember have integrated t he ef- fects of the 3 variables i nto a "rnoc5-e l" : ... patrilocal residen ce i s favored by the pre s e nce of at least s ome internal warfare (that is , warfare within the s ociety), whether or not such war fare i nterferes with a patridominant division of l abor ; and matrilocal residence is favored by pure l y external warfare if such warfare compe l s t he divis ion of labor t o become matridomi nant . (1 971 : 503 ; italics added ) . Note that the Embers ' use of the phrase "favored by" i s a translation of coefficients i nto the i dea that warfare is one i ndependent variable , patridominant l abor and patrilocal res~ i dence are two other independent vari able s , and ma tridominant labor and matrilo cal residence are variables dependent upon warfare . There is some inconsistence : "internal warfare" and 11 external warfare " are variables dependent upon the form of residence . So i s warfare independent or dependent ? As a change i n point of view from the s tri ctly politi- ca l toward broader i nfluences of culture upon the expression of human aggression , Russell (1 972) performs the purely i nduc- tive technique of factor analysi s upon a l arge group of char- acterist ics drawn from Textor t s A Cross-Cultural Summary (1 967 ) that have to do, i n Russell ' s estimation , with aggres- sion . His t a s k is "to i sola te the primary groups or patterns of cultural characteristics re l ated to warfare " (1 972 : 279 ), a ctually to produce a psychocul tural expl anation of war . He ha s ma de up a matri x of Pbi coefficients of the corre l a tions 240 betueen Textor ' s measures--bellicosity, emphas is on military glory, emphasis on killing and torturi ng enemies , amount of war f are--and another group of measures which , although they do not have a direct relationship with the core warfare vari- ables , do have strong re l at ionships with other variables that are re l ated to the warfare variabl es or that may indicate new leads , e . g ., 5 geographic areas, which subsequent l y do not become part of any factor . Russell uses 78 vari abl es alto- gether , f or 400 culture s . What he has found are clumps of variables sticking to each other . . Military glory, belli- cosity , torturing and killing of t he enemy , and the amount of uarfare clump , along with murder , assault, and theft . Thus, a ll measures of every type of aggression that have been examined l oad on thi s one factor , and apparently the more violent the aggression the stronger the loading . This factor , then appears to represent not just warfare but a l l forms of cultural aggressi on , i nternal as well as exte rnal to a society , which means in turn that a ll forms of aggression tend to be strongly re lated to each other . Evidently cultur es tend to vary not only i n their ten- dency to be warlike but also i n their general level of hostility, and this hostility will take any and all availabl e forms (Russell 1972 : 29 1) . Through the shift ing and sifting of factor analysis , Russell has als o found that narci ssism , indicated by "boastful ness ," "se nsitivity to insult ," and "invidious display of weal th ," i s close ly rela ted to warfare (1972 : 292 ). He also i so l ates a third clump of characteristics- -achievement mot ivation : "pressure to a chieve ," "anxiety over not achieving ," "religi.on supports a chievement ," "invidious displ ay of wealth ," and "existence of entrepreneurs ," commenting that 241 entrepreneurial activity may be a pproved and sublimated cul- tura l hostility , i . e ., a mirror image of theft . Clinging to the achi evement clump, Russell has found that only 2 of 12 anxiety measures (ironically, "aggression anxiety " and "anxiety about being self-reliant 11 [ as contrasted to anxiety about not being self-reliant]) were not positively related to warfare . Thi s is strong evidence of the great amount of under- lying anxiety and i nsecurity found i n warlike cultures . Thus , it is possible that high levels of hostili ty repre- sent as much attempts to compensate or defend against feelings of i nse curity and anxiety as an expressi bn of i ntrinsic aggression ; of cours e, both could be involved . I n any case , this factor indi cates that membe rs of a war- like culture are not only more hostile but a l so more narcis s istic and insecure than members of peaceful so ciety (1972 : 295 ). Russell ' s other psychocultural findings that have pos- itive relationships with the warfare factor are "infant aggression satisfaction ," "extramarital sexual relations are pu.I1i shed , 11 "premarital sexual relations are punished , 11 and " sex anxiety i n adult i s high" (1972 : 296 ). In summary , when all these result s are examined together, a rather consistent picture of the psychocultural aspects of warlike cultures appears . It i s evident that warlike- ness is only one form that hostility i n a culture may take , and that the level of hostility i n a culture varies as a whole from culture to culture . This host ility may appear in the form of bellicosity , personal crime, theft , or emphasis on achievement , weal th , and entrepreneurial activities . In regard to dynamics , the l eve l of host il- ity is posit i vely related to the amount of restrictive - ness of punitiveness that the culture places on i ts mem- bers at all age levels. In adolescence and adulthood , the punishment of extramarita l and premarital sexual relations is particularly evident . The result of this punitiveness is evident not only i n cultural hostility but a l so i n a deep sense of anxiety and i nse curity, which appears in narcissi sti c att itudes , such as boasting and 242 sensitivity to insult or an exhibit ionistic di spl ay of weal th , and a l s o i n the many measures of anxiety used i n this study . . .. The members of such cultures are certai nl y neither happy nor "mentally healthy ." Thus , the ramificat ions of warlikeness extend far beyond be llicose behavior itself . However, it should be noted that this effect i s a genera l t rend among cultures and that any s i ngle culture may not exhibit this pattern (Russell 1972 : 29 7) . Russell goes on to a r gue that formal characteristics of cul- ture , such as descent and complexity , are probably only weakly related to warfare , and that t he clusterings of psy- chocultural variables are not only strongl y related to war ·but are a l so strongly r e l at ed t o each othe r and t hus are a manifestation of universal features . Therefore , while materi a list i nterpretat ions of di fferences among cultures and of hist ory are i mportant, fo r some phenomena such as warfare the psycho cultural interpretation is much more i mportant -- political and economic causes "rest on a general level of host ility and anxiety . A high level of hostility probably potent i ates certai n political a ctions as well as maki ng other actions i mposs ible '' (1 972 : 304 ). Moreover , if psychocultural variables are primary , not formal or materia l ones , 11 ••• highly c omplex cultures could be as peaceful as simple cu l - tures if they had the right psycho cultural patterns " (1 972 : 303 ) . The "right " pa tterns would be sat i sfact i on of needs and reduction of social anxi ety . I n t he face of Naroll ' s ( 1966 ; 1969) and 0t terbe i n 1 s findings (1970) t hat warli ke culture s tend to expand and the i mplicat ion t hat therefore 243 chang i ng warlike attitude s woul d amount to "psychologica l disarmament ," Russell conc l udes with the observation that We s tern societ ie s can afford c onsiderable change : '' ... with the high level of hostili ty that now exist s i n this socie ty a great amount of change could occur without placing us i n danger of being overrun ; rather it would probably enabl e us to take a more posit i ve role in creat i ng world pea ce" (1 972 : 306- 307). While Russell ' s claims are a refreshi ng change from t hos e of the kill-or-be- ki lled school , one i s still left with the problem of inherent conditions--imperatives of t he psyche this time . Neverthe l ess , as a possible insight into why pat - terns of aggression are perpetuated generation after genera- tion and may fall upon especially re ceptive ground i n diffu- sion , an explanat ion formulated in terms of deprivation and anxi ety , despi te persistent vagueness and circular ity, speaks i n more life-like terms despite its gene tic relationshi p to the frustration- aggression hypothesis . Russell i ncludes 5 geographical areas in his variable s --Africa , East Eurasia , Pacific , North America , South America--but only ment ions geography once in his findings i n a passing comment about possible l oading in Africa . History and diffusion once again are dropped out , whi ch is to be expected , I suppose , in most psychological approaches . I also woul d like to note that insecurity and anxiety may not generate hostility and thereby warlikeness but can arise out of trying to live up to the 244 cultural i deal s of a warrior soci ety . Yanomamo warri ers com- monly get stomach aches and sore feet on the way to a rai d , whi ch excuses them from further a ction . Modern businessmen wear out their hearts , stomachs , and nerves try i ug to conform to cultural s t andards of a cceptable behavi or , an achievement that most can state they want . Sipes (1 973 ) t ests 2 fami liar compet i ng explanat ions for r e lationships a mong aggression , sports , and war--the Drive Di s charge Model and t he Cul ture Pattern Model . The former c laims that aggression , whether i nnate or a cqui r ed , builds up i n an i ndivi dual and can be drai ned off t hrough sports , as an a lternative to war ( cathartic dis charge ; func- tional or mora l equivalents ) . Thus , those societ ies with more war i:,-;ill have less combat i ve sport s activity , and t hose wi th l ess will have greater sport s a ctivity . The second mocel claims that aggressive behavior i s primarily l earned and i s therefore cultural, and that since behavior and values associated with war are a l so associated with sport s , the 2 phenomena reinforce rather than neutralize ea ch othe r and their relat i ve presen ce or absence will be found to vary t ogether (cultural cons i stency ; t hemes ; configurations ) (Sipes 1973 : 64- 67) . Sipes se l ects 10 relative l y warlike societ ie s from 0tterbei n 1 s 1968 sampl e that had adequate data on sports , but since he is unable to mat ch t hem with 10 relatively peaceful societ i es from the same source , because 0t terbein rs sarr..ple 245 could supply only 4 , he a dds 6 of his own choice . The 10 societies having frequent or continual war that Sipes uses are Tibet , Thai , Serna Naga , Ila , Cornox , Aztec , Tehuelche , Abipon , Timbira , and Mundurucu . The 10 peaceful socie t ies are the Toda , Dorobo , Copper Eski mo , Tikopia , Semang, Bhi l , Hutter- ite , Lapp , Kung Bushman , and Naskapi- Montagnais . Sipes ' results for the 20 pri mit i ve so ciet ie s are : Combative Sports Yes No Yes 9 1 10 Warl ike No 2 8 10 1 1 9 ~ = . 6035 FE< . 0028 (1 973 : 71 ). Obviously , he concludes that the Culture Pattern Model is valid while the Dri ve Discharge Mode l i s not ( 1973 :71). Sipes performs a se cond test of the 2 competing theo- ries upon a single case over time , the Uni ted Stat e s of America from World War I I t hrough the Korean Confli ct and the I ndochina War . He has broken down the dependent variable of sports i nto 20 spectator/participant and combative / non- combative types , select i ng 4 as type- specimens : f ootball (combat ive- spectator) , hunting ( combative- participant ), base- ball (non- combative- spectator) , and race betting (non- combat ive-participant ) . Using quantitative data as indi ces 246 of re l ative intere s t in t he sports--hunting licenses i ssued , horse racing revenue, number of spe ctators at National League baseball games a nd National Football League game s - - and inter- corre l a ting and graphing the s e wi th the percentage of adult male s in the military from 1920 to 1970 , Sipes agai I'- ha s found support fo r the Culture Pattern Model . Hunt i ng , foot - ball , and race betting show a continuous , steep , upward trend over the half- century , while baseball (the non-combat ive , :non-participant sport) has experienced an overall slight decline in interest , with sharp drops during all 3 wars . As a fina l observat ion, Sipe s no tes t hat since the Drive Discharge Model i s so much a part of We s tern science, there should be further te sting of the drive versus the pat- terning mode l on other expressions of violence , such as "sui- cide , murder , punishment of deviants , drug use , physical assault on family or other community members, gossip, psycho- genic illnesses , and ma levolent magic " (1 973 : 80 ). Despite t he methodological rashness of using a sample of merely 20 s ocietie s--a practice apparently not di s couraged by Naroll and Otterbein, whom Sipes credits with crit ica l revi ew (1973 : 80) --I am impressed by Sipes ' wi llingness to test a dearly held belief and broadcast it s failing to the world . Given t he limi ted variables , there should not be much difficulty in expanding the sample size to one more reliable . I hope that Sipes does just tha t . 247 My pre liminary cross- cultural warfare study in 1968 (Nammour n . d .) is concerned with the participation of women i n warfare , testing 7 specifi c hypotheses on an availability sample of 35 societies sprinkled around the world : New World 13 , Africa and the Circum- Mediterranean 11, Eastern Europe and the Insular Pacific 11. Ea ch of these cu l tures belongs to only 1 of Murdock 's (1967) culture clusters . Some control for diffusion i s achieved by using Murdock ' s criterion that cultures be separat ed by 3° of latitude or longitude i n the tropical zone, 4-5° in the temperate zone 35° from the eq_uator , and 6° in the frigid zone . The hypotheses t ested are : (1) The participation of women in military activities is associated wi th the frequency of warfare . (2) The parti- cipation of women is more strongly associated with military expectations of revenge, defense , and prestige than wi t h expectations of territorial gain . (3) The leve l of partici- pation of women i n warfare is dependent upon the political c omplexity of a society . The relationship i s obve r se : if political complexity i s high , participation of women will be low . (4) The participat ion of women i n warfare de creases as the dominant subsistence activity be comes more complex. (5) If women are domi nant i n the division of labor in the domi- nant subsistence a ctivity, but war i s fre quent, female par- ticipation will be low because they must devote their ener- gies to subsistence activity in orde r to support the military compl ex . (6) Where warfare is frequent , residence will be 248 patrilocal , but where participation of women i s high , resi- dence will not be patr-i local , on the assumption t hat a woman residing with her consanguineal kin will be more i nvo l ved i n warring act i vit ie s than a woman residing with her husband ' s family . (7) Since des cent has been demonstrated to succeed residence ( Driver and Massey 1957) , t he relationships among descent , frequency of warfare , and participation of women are expected to show stronger relationships between respec t i ve attributes but i n the same direction as correlat io ns for res- i dence . Only hypotheses (1), (2), and (4) produced _statisti- cally significant results , despite elaborate partialling of variables . I nc l uding tendencie s indicated by Yule ' s Q and the phi coefficients , although they are not significant , in 1968 I i nterpre te d the intercorrelations as hinting at the follow- i ng relationships. If a tribe i s i nvolved in frequent war- fare, women will participate v i gorously , especially in r itual activity . Women participate i n war for prestige , but Lot for revenge, defense , and plunder . When the l evel of politi ca l complexity is hi gh , i . e . , when 2 or more jurisdictional l evels beyond the local level exi s t , women do not participate in warfare . Women will be most i nvolved in warfare in soci- eties having hunting and fishing e conomies , for which the highest military expectation is prestige . While warfare may be frequent in t hose societies with matri- dorr.inant l abor , women do not seem to part icipate be cause they are active in providing enough to eat through farming whi le the men are 249 off to war . Frequency of war is asso ciated with patrilocal residence but more strongl y associated wi th patrilineal des cent . This hint i s i nteresting in light of Otterbein rs f i ndings ( 1965 : 1968 ) that patrilocality i s not an i ndi cat or t hat men will a ct in concert, while polygyny is . Perhaps mere r es i dence i s not enough ; jural cla i ms fo r cooperat ion , whether t hrough patrilineality or po l ygyny , must ex i s t . Wa r does no t occur often wi th matri l ocal resi den ce and l ess often with mat r ilineal descent . Part icipat ion of women in war i s l ess a ct i ve wi th patrilocal resi dence than wi th pat rili nea l des cent and does not oc cur with mat riloca l resi dence or matrilineal des cent . Methodol ogically , t h i s work i s fatally weak i n s ample size , and any f i ndings are offered only t ent at ively and a s t hings t o l ook f or usi ng a l a r ger sample . Other t han sampl e , ~he me chani cs of my or i ginal warf are study are s ound . This concl udes my presentat i on and di s cussi on of pre ceder..ts i n quant itative warfare research i n ant hropo logy . The next section presents the coding and stat i s tica l pr o- cedures that I have used i n the quar.t i tat i ve research at hand . 250 Coding and Statistica l Procedures ·1 • The data . I n my research effort I used 2 ki nds of dat a : material analyzed and compiled i n the Ethnographi c Atlas (Murdo ck 1967 ) a nd i ts successor i n refinement for the Stand- ard Sample , the "Cros s - Cultural Co des " (Murdo ck and Darrow 1970 ; Murdo ck and Wilson 1972 ; Tuden and Marshall 1972) , and publi shed ethnographic s ource s that I have coded for the war- f are vari abl es . Appendix E contains the bibliographi c refer- ences used , listed separately by society . Reliability of the published codes I have assumed to be adequate , and the newer "Cross- Cultural Codes " i ncl ude i nformat i on on reliability che cks . I n any kind of categori zing there is bound to be some disagreement between c oders and disagreement by anthro- pologi s ts using the co des with the decisions made by their architects . At the same time, occasional di sagreement does not seem to be adequate grounds fo r dismissing the codes out of hand as i nevitab l y wrong and therefore useless . I n my experience , it seems t hat some of tho se who at t a ck quant i f i - cat i on a r dent ly are themselves engaged in c ontinuous cross - cultural comparison, a l beit on a non-mathemat ica l basis . I n gatheri ng and coding i nf ormat ion on warfare , t wo other per- s ons a lso have done s ome codi ng , and a portion of their work I repeated i ndependently as an inforraal reliability che ck . It s i nformality exists be cause I have not mathematically 25 1 measure d the percentages of disagreement between myself a nd other code r s . The che ck doe s have a formal aspe ct i n that I s crupulous l y di d not l ook at others ' de cisions before I made my own . The maj or difference between the other coders ' work and my own i s that I have fel t compe lled i n most cases to read comprehensively on t he culture before making codi ng decisions--in a sense putting the culture back together before I coul d take it apart , again . My note-taking to support coding decisions tended to be more copious . Doing such con- textual work does increase one ' s i r rit at ion with the injusti- ces of classificat ion; the re i s s o much that must be cut off around the e dges . Occasionally I have us ed ano ther coding technique , one used by Ember and Embe r (1971), in which I worked together with a se cond coder maki ng cooperative deci- sions . The cultural representative s of Japan and China are 2 example s of s ocie t ie s dealt with in this manner . Below I present each var i able , its attributes , and the codi ng procedure . 2 . Stat i stical manipul at i on of the data . When I wrot e the proposal for t hi s proje ct , I had planned to use the Stanford Statistical Package (Nie , Bent , and Hull 1972 ) for the computer work on the i ntercorrela- t i ons , but at t he time California State Uni versity, Sacra- me nto , unlike the University of Oregon , did not own the compiler for that program , and none of their exist ing canned 252 programs sat i sfied my needs . Therefore , I had a program written to intercorrelate every variable wi th every other variable i n fourfold t able s and to perform 8 statistics on each correlat ion , to exhaust the mathematical possibilit i es of a limited format using qualitative variables . The statisti cs include Yule ' s Q, phi , chi- square , Pearson ' s C, Goodman and Kruskal ' s Tau on col umns and rows , and Goodman and Kruskal 1 s Lambda on columns and rows . As gross indicators of geographical variation in corre l ation coefficient s , I used Pearson ' s ~ - Yule ' s Q is a measure of 1-way association and has virtually no utility by i tself , but when used with phi, which is a measure of 2- way asso ciation , t he coefficients complement each other and can be used only on 2 x 2 tables . For instance , if 2- way association is ve ry weak in a part icul ar cas e , there may still be a strong 1- way association between variables which c ertainly should be sal- vaged and may be quite i mportant i n dete cting the di re c t ion of at tra ction between variables . Chi -square , of course, i s t he measure used to derive probabi lity val ues . Since chi- square raw values are not comparable , Pearson 1 s C is a normed va lue derived from chi- square that i s comparable to other Cs . Using fourfold tables , phi a nd Care i dentical or close i n value for lower coefficients , but s ince C can reach unity and phi cannot , higher va lue s manifest i ncreasing discrepancy . For i nstance , i n a perfect correlat ion , C will be 1 . 0 while phi will be . 707 , with e i ther a pos itive or negative sign. 253 Carrying a s i gn i s the virtue that phi has to counterbalance C' s abi lity to reach uni t y but wi thout a s i gn . I f I reli ed upon Cs alone , I would not know whether a relationshi p was a perfe ct posit i ve correlation or a perfect negative corre - l ation . Lambda and tau are measures of association that give predictive values : how often does knowl edge of the i ndepend- e nt variable allow one to predict its association with a par- ticular dependent variabl e . The beauty of both the se meas- ures i s that they are dire ctional . Knowle dge of variable A . may not gi ve any capac ity to predict A' s associat ion with B, but knowle dge of B may a llow one to predi ct t he pre s en ce of A to a specified degree . While l ambda and t au do t he s ame sort of thing , they are mathemat ica lly de rived i n different ways . By and l arge i n my work lambda has been more u seful than t a u , but t here are i nstances i n which lambda may be 0 but tau may be greate r than 0. By u s i ng both measures , I hoped to extract a ll possible i nformation from the relation- ships between qualitative vari a ble s . Since Q and phi are not predict i ve measures , lambda and tau are powerful tools . It i s poss i bl e to have high phi value s and h i ghly s i gnificant chi- square value s but no predic t i ve capability at all--lambda and tau may be 0 . I n such a case , the non-predictive meas - ures of association simpl y summari ze how often i n one t s sample the vari ables occur together but give no i nsig~1t into whether or not the variables mathemat ically generate indi- cat ions of an inherent re l at i onship . The absence of such 254 indications severely limits the interpretation of stat i st ical result s , especially for tho se workers i nclined toward universal theorizing . The computer program has performed one last service. Ea ch society i n the Standard Sample has an identity number . Each member of every cell in the table s is identified by its number, so t hat I not only have cell frequencies but also know who i s where. This i s especially i mportant for my basic hypothesis, of significant geographical variation i n the configurations of the variables . Each variabl e has been i ntercorrelated with every other variable i n the interests of program efficiency , first for all the societies within each of the 6 geographical regions and then for all of them together forming the world- wide sample . Fuither statistical analysis of relationshi ps I have performed by hand . The interpretation of all the mathe - matical bits and pieces is by i nspe ction . This task can be likened to creat ing a 4- di mensional scene (iLcludi ng time ) out of 2-dimensional puzzle pieces . The Variables Because I have used such a large number of variables in my research , the most straightforward way to present them is to li s t each one , giving for each the source of the data, bow it has been coded , and how I have categorized the attri- butes for statistical manipulat ion . The first group of 255 variables are familiar sociocultural ones that I thought per- tinent to warfare at the time I selected them . The second group consi sts of variable s dealing with the practice of and attitudes toward warfare . Each variable has a nwnber assigned fo r the computer program and retained for all tables i nclude d here. A separate nwnber is used for each variat i on of a basic variable . Refer to Appendix B for the code sheets and to Appendix C for the variable codings for each society i n the sample . 1 . Regional identification . I have used the identifica- tions from the Ethnographi c Atlas (Murdock 1967 ) , hereafter referred to as the EA , of Africa , Circum- Mediterranean , East Eurasia , I nsular Pacif ic , North America , and South Ameri ca without breaki ng these regions down into fine r units, since the Ns would then be too small for any s tatistical analysis . In t he computer run on the worl d sample, I have treated each geographical region as a variable , to be i ntercorrela te d with all other variable s but not with ea ch other . 2 . Polygyny . This form of the family has been code d present i f it has an incidence of at least 20 per cent for a given society and absent if less . Origina lly I use d the codings of the EA but then revised them to conform to the revisions of "Cross-Cultural Co des 3" (Murdock and Wilson 1 972) . 3 , 4 . Mari tal residence . The attributes of thi s variable have been scaled 2 ways and each correlated with all the 256 other variables . The first dichotomizes resi dence i nto patri l ocal or virilocal on one hand and matri local , avuncu- local , or uxorilocal on the other . The second dichotomizes r esi dence into patrilocal , vi rilocal , matri local , avuncu- loca l, or uxorical, i n contrast to ambi local or neolocal . The s ource f or t he codi ngs i s the EA. 5. Communi t y organi zation . Attri bute s fo r t his vari abl e as coded i n t he EA have been lumped , pe rhaps awkwardly , i nt o dichotomi es i dent i f i ed as endogamy prevalent ( de.mes , seg- mented communi t i es wi t hout l ocal exogamy , agamous communi- t i es) and exogamy preva l ent ( exogamous communi t i es , segmented communi t i es with lo cal exogamy , clan-communi t i es ) . The source i s the EA , although CCC3 has a be t ter variabl e , nintercommunity marriage ," that more suitably des cr i bes t he se marriage patterns . My l ater judgment i s t hat I shoul d not have used the EA data . I attempted t o use t he CCC 3 mat er i a l to revise the EA codings and was somewhat successful , but t he 2 vari ables are not t he same and the result is patchwork . 6 , 7 , 8 . Sett l ement pattern . The attributes of the va r i- ables have been scaled 3 ways and each intercorre l ated wi th all of the other variabl es . The f i rst lumps nomadic bands, semi nomadic communities , and semisedentary communi t i es i n contrast t o &11 other types ; the second l umps compac t but i mpermanent settlements , neighborhoods of disperse d family homesteads , and separated hamlets ; the thi rd does the same wi th compact , relatively permanent settlements and c omplex 257 sett l ements . The source of data i s the EA . 9 , 10 , 11 . Mean s i ze of l ocal community . This variable is scaled 3 ways , opposing each focal cluster of a ttributes to a ll the others i n the di chot omy . The firs t l umps corrunu- nities with fewer t han 50 persons , 50-99 persons , 100-199 persons , and 200- 399 persons . The s econd s ifts out the small est communities--fewer than 50 persons and 50- 99 persons--while t he third i n turn pull s communities ou t having 100- 399 pe r sons . I n all 3 scalings , t he attributes consist - ent ly placed i n oppos i tion i n the dichotomy are communities of 400-1 000 persons , 1000 wi thout any town of more than 5000, 1 or more towns of 5000- 50 ,000 , and 1 or more cities of more than 50 , 000 persons . I have pl a ced so much emphasi s on the smalle r groups , trying to find some population floor beneath ..,;hich a community canno t support i ntensive warfare . The code source was initia l ly the EA , corre cte d by t he r evisiorrs in c ommuni ty size co ded i n CCC 3 (Murdo ck and Wils on 1972) . Population density per square mi le . Fortuna tely , "Cross- Cultural Code 3 11 (Murdock and Wilson 1972 ) a l so i ncludes coding for this variabl e . Codi ng it i ndependently would have been diffi cult . Murdo ck and Wilson concentrated on " ... the density of populat ion in the area exploited or controlled by the fo cal or typical communi ty 11 ( 1972 : 257 ) . The authors developed 7 attributes , which I have scaled i n several ways : each one against a ll of the others , and t hen di chotomie s that su ccessi ve l y move one attribute at a time from one side to 258 the other . This variable , vital as it is , was not i nc l uded i n the f ull computer run . Preliminary hand- computations are di s cussed in Chapter III . 12 , 15 , 14 . Political complexity . This variable is based upon Murdock ' s variable "Jurisdi ctional Hierarchy" in the EA , which i s coded for both the number of local jurisdictional levels and the number beyond t he local communi ty . "Cross- Cultural Co des 3 and 4" also i nclude variables on political organization , but I found them to be either too detailed and tedious or too broad to be of use to me at this time . I did use the CCCs as a check against my interpretation of the original rough EA classificat i ons . This variabl e remains a crude indicator , possibly because political organization and the distribution of power exists in myriad forms . Local com- plexity is dichotomized i nto 0- 2 against 3 or 4 levels . Com- plexity beyond the l ocal community is scal ed 2 ways : 2- 4 levels in contrast to O or 1 level , and 1-4 levels against the absence of any superordinate jurisdicti on . In the termi- nology of CCC 4 (Tuden and Marshall 1972) , from 2 to 4 juris- dictional levels would correspond to petty paramount chief- doms , small states , and l arge states. 15 . Subsistence . The dominant subsistence activity of each society has been intercorrelated 2 ways : a dichotomy between food- collectors (gathering , fishing , hunting) and food- producers (pastorali sm , i ~cipi ent agriculture , extensive agriculture , intensive agri culture) , and each subsistence 259 te chnique taken separat e l y but i nt ercorrelated with only a few se l ected warfare variables . The source for the subsist- ence coding i s primarily the EA, supplemented by "Cross- Cultural Codes 1 11 (Murdo ck and Morrow 1970) for those Stand- ard Sample societies not i ncluded i n the EA. 16, 17. Descent . The attributes fo r this variable have been scaled 2 ways . The first const i tutes a dichotomy between patrilineality and mat rilineality only, the second po se s unilineal aga i ns t duolateral, bilateral, and ambi - lineal organization . The codi ng source is the EA . 18 . Division of l abor . With this variable I wanted to test for any relationship between the warfare variabl es and which sex was doi ng mos t of the work i n the dominant subsist- eI'-ce a c tivity , followi ng up an i dea by Ember and Ember (1 971 ). The at tributes of the vari able are dichotomi zed only 1 way : males a lone or almost alone and males appreciably more con- t ras ted to a lumping of equa l part icipation , females appreci- ably more , female s alone or a l most alone , a nd sex i rrelevant . The source of the codi ngs i s t he EA . I have made a mistake i n not i ncluding as variables the second most i mportant sub- s i stence a ctivity of a society and which sex does most of the work i n that a ctivity . I am suspicious that the assumption that if men do most of t he work i n an a ctivi ty t herefore it must be the domi nant one ha s i Lf luence d judgments in co di ng t hese va riables . If this i s so , then possibly even with such data the effect of women i n subsi stence may be t eased out of 260 coding activities judged no t to be the most important . Whi le t he Cros s - Cultural Codes i ncluQe refi ned coding on subsist - ence , they do not i nclude division of labor for any act i vity . 19 . Class stratification . The attributes of this vari - able from the EA a r e s i mpl y dichot omi ze d i nto one cluster that i ncludes an elite based on control of resources , an he reditary ari stocracy , and s oci a l classes cont r asted wi t h a s econd made up of the absence of class stratification among freemen or di s tinct i ons of weal th only . The source i s the EA . 20 . Succession of local headman . The attributes of this variable are dichot omi zed i nto heredi tary succession , r egard- less of the rule of inheri tance , and nonhereditary succes- sion, whether by appo i ntment , seniori ty , i nfluence , ele c t ion , or i nformal consensus . The source i s t he EA. The remaining variables are desi gned to deal wi th the i ss~e at hand , warfare i n human soci ety . I shall cont i nue to l i st them, givi ng the di chotomies of the attri butes , t he rational e for i nclusi on of the variable , and s ources f or tho se used elsewhere . Following the pre cedent set by Embe r and Ember (197 1) , i f warfare does not exi st at the time pi n- poi nted by Murdock and Whi te (1969 ) , but it has exi sted within 50 years prior to that date , I have coded for i t as t hough i t were still going on . I n such cases what usually has happened is European contact and picification . The effect of this practice i s to recategorize some societies i n 26 1 the tiff.e slots i dentified by Murdock and White, i n particular putting ones pinpointed i n the first half of the 20th century back into the late 19th century . 21 . Frequency of war between political communit i es of the same cultural unit . This i s what Otterbein calls "internal war ." Otterbein (1 970 : 3 ) uses Naroll ' s defi ni tion of a " ter- ritorial team ," but renames it a politica l community, as a "group of people whose membership is defined i n terms of occupancy of a common territory and who have an official with the speci a l function of announcing group decisions--a func- tion exercised at least once a year " (Naroll 1964 : 286) . Otterbein does not use Naroll' s compl ementary concept of a "cultuni t " but chooses instead Malinowski ' s ( 1941) concept of cultural uni t . Cont i guous polit i cal communities which are culturally similar compri se a cultural unit . I n most i nstances , t he cultural unit is the same as a society , which i s the uni t used i n the universe from whi ch the sample is drawn . Although a s i ngle political community may be cotermi nous wi th a cul tural unit , a cultural unit usually consists of more than one political community (Otterbein 1970 : 3 ). These definitions are largely academic, however , because Otterbe i n uses cu l tural units as defined by Murdock in the Ethnographic Atlas , and I have used t he t arget communi t i es specified by Murdock and White (1969) for the Standard Sample . Some estimate of occurrence is obvi ousl y basic i n war- fare studies . The attributes are continual (perpetual , con- stant), frequent (common , intensive) , and infrequent or never 262 ( occasional , sporadic , rare) . The dichotomy is between con- tinual and frequent war , and infrequent war . The form of this variable close ly follows that of Otterbein (1 970 : 143) . Selecting the most appropriate coding slot i s based upon terms used by the ethnographers themselves (e . g . , "cont i n- ual," "common") or , where contextual des cription does not appear to support the ethnographer rs adjective , a revised assessment by the coder supported by notes on the pertinent i nformation . Thi s vari able does not distinguish between the target community and other communities as to who is attacking and who is be ing attacked . Thi s i s a weakness in the vari- able , but i t probably cannot be helped be cause the target societies in the Standard Sampl e may be either a s i ngle c om- munity or the whole culture , whose member communities one discovers do fight with each other . The di stinction between feud and warfare wi thin a cultural unit i s of course observed , usually relying upon the identifi cation of the eth- nographer . Doubtless many cases of such i dent i fication will be disputed by other anthropologists , leading to all the def i ni tional contortions discussed at the beginning of Chapter I. 22 . Frequency of attack by political communit i es of the cultural unit upon members of other cultural units . This variable and the next are what Otterbein ca lls "external war ." The same indicators for coding decisions are used as in the previous variable and. the dicbotomie s remain the same , 263 but t he task of distinguishing between feud and war is obvi- ously much easi er . Tbe l iterature often fails at this point , however , by mentioning the frequency of warfare between spe - cifi c parties but i gnoring the di fferer..ce between the i nci - dence of attacki ng and being attacked . 23 . Frequency of being attacked by non- members of the cultural unit . The dis cuss i on of t he previous vari abl e i s appli cabl e here , also . Bo t h variables closely f ollow Ot terbein ( 1970 : 143- 44 ). 24 . Mobili zation . Otterbein ' s (1 970 : 144 ) vari abl e con- s i sts of 4 a t tri butes , making fine distinct i ons between non- p:cofessi onals and professionals ,, apd combi nations t here of . I emphasize the organizati onal units and base the dichotomy on the attributes of age- grades , military soci eties , and stand- i ng armies , as opposed to organizat i on based i nstead on friendshi p or kinship , or the absence of organi zation alto- gethe r . This vari abl e has been one of t he easier ones to code , with usual ly adequate ethnographic descript i on . I am interested i n the form of mobili zat i on not as a military t a c- tic , which i s Otterbe i n ' s concern , but as the basis for the armed group with the common goal , that i s , one of t he 3 neces sary conditions for warfare . Furthermore , I am i nter- . e sted i n organi zational principles that bring males 1 together on a regular basis , especially where mili tari sm may be a principl e upon which the organization is built , which in turn may perpetuate military attitudes and exercises . The 264 difference between the 2 types of grouping i s se l f - evident : age - grades , and particularly military so cieties and standi ng armi es , exist i n primary relationshi p to military a ctivities ; friendshi p a nd kinship groups may i nclude military activity as merely another of a variety of activities and obligations . 25 . Decision to go to war . This variable i s i denti ca l with that of Otterbein ( 1970 : 144 ) and consists of only 2 attributes : the decision to go to war is made by an official or council of t he .political community, or anybody can decide on his own without reference to a higher authority . Again, Otterbein include s thi s variabl e as a military t act ic--an officia l de cis ion is more highly evolved socioculturally than an individual one . My concern is entirely different . For i nstance , a s I pointed out i n my dis cuss ion of Otterbein ' s work , he sees higher political authority as bringing i nherent i ndi vidual violence under control . I n contrapos ition , I see decis ion- making authority vested i n the few as likely to affect many more persons with violence t han if they acte d as i ndivi d.ual s . Di ssent i ng opinions . I had hoped to be able to gather ethnographic material on how members of a society receive dissent i ng opinions i n regard to war , i . e . , protesters , paci- fists, di p.lomats , and if any attempts are made to stop one ' s group fro m going to war . Thi s variab l e obvious.ly i s based upon our own polit ica l history . Since ours is only 1 culture among hundreds , I wi shed to di scover if i n other cu.ltures 265 attempts were made to dissuade fellow citizens from waging war , particularly an offens ive one . Sadly , the information simply does not exist except i n a few superl at ive cases . I can conclude only that the ethnographers have not asked the quest i on . This variable i s not iriclud.ed i n the quant i tative analysis . 26 . The beginni ng of war. This i s another variable tha t I have borrowed from Otte rbein ( 1970 : 144 ) . While his is again tactical , my concern is with the circumstances under which warfare i s carried out openly and with s ome measure of agreement between the parties . The 3 att ributes of the vari- able are dichot omi zed by dist i nguishing between war begun by announcement or mutual agreement and war b egun by surprise 2.ttack . 27 . The ending of war . I have borrowed t his variable f~om Otterbein (1 970 : 144) , with very slight modification . The dichot omy contrasts conclusion by negotiat i on with con- clusion by simply stoppi ng or not concludi ng at all. 28 . Peace ceremony . I think that a peace ceremony usu- ally gives overt sociocultural recognition of the co sts of war and the intentions of participants in the future . While it may be a sham , the existence of the not ion of fraud i n such an event also attests to its positive importance . Occasionally in the literature one comes a cross comments by anthropol ogists to the effect that a peopl e engage in perpet- ual war because they have no institutional means to end it . 266 Negotiation , where it exists , is a means to such an end , and the peace ceremony i s the ritual that sanctions the end . It is not necessarily the case , however , that negotiation i nevitably i nclu ~es a ceremony . This vari ab le has been coded simply as e i ther present or absent . 29 , 30 ; 40- 47; 49- 52 . Military expectations . Recall that the concept of military expectations was originated by Naroll (1 966 ), who argues that while i t is extreme ly tenuous to talk about the objective cause of a particular war , one can readily colle ct information from the a ctors as to t heir reasons for go i ng to war . Further recall my repeated argu- ment that objective cause does no t tell us much anyhow , because people act i n a ccordance with what they believe to be the case--that the military expectations of the people are not second-best data but the data we should concern ou_rselves with first and foremost . This variable is i nit i a lly based upon Naroll , with modifications by Otterbein (1970), and augmented by 1 new attribute . Naroll ' s original attributes , lis t ed in order of ri sing expe ctations , were : Revenge and defense . Tribes fight to gain satisfaction for i njuries (like murder or witchcraft spe lls), or to expel a foe from their territory . All warring tribes have this expectation. Revenge , defense , and plunder . Tribes may also fight .. • for booty of some economic value--catt l e , wives , s l aves , land , cannibal victims ( ... consumed for food ) . Revenge , defense , plunder , and prestige . I n addition to both expectations ab ove, warriors wilJ go to battle to 267 prove their mi l itary prowess--for instance , to acqa ire scalps or victims for ceremonial cannibalism . Revenge , defense , pl under , prestige , and political con- trol . The i ncorporation of the defeated enemy into the poli t ica l system of the victor becomes an additional goal of warfare (Naroll 1966 : 17- 18 ) . I used these attributes i n an earlier study (Nammour n . d . ) wi th one modification : I judged the l ast attri bute , politi- cal cont rol, to be present i f a t r i be went to war for terri- torial c ontrol. For example , the Nootka (who are not i ncluded in the Standard Sample) go to war to wrest fishing and gathering territory from other peoples , but they also practice war of attritior. tempered by slave taking . I n any case , at that time , if te rritory were mentioned by the eth- nographers as a goal in warfare, I consi dered the fourth attribute to be present . Nevertheless , contro l of territ ory and its natural resource s and i ncorporat ion of a defeated people i nto the vi ctor ' s political system are no t necessarily the same thing . Subsequently , I have borrowed Ot terbe i n ' s (1970 : 146) attribute of land--fields , hunting territories , fishing terr itories, pastures--as a distinct expectat i on from e i ther subjugation of a people or plunder . Furthermore , pl under and tribute are diffe rent sorts of econoI:1i c gain . The Aztecs exacted tribute fro□ defeated ci ties and t hus dom- inate d surrounding peoples , but they did not hol d domi nion over them . Their colle ct i on of tribute i s an entire l y di f - ferent political act from others 1 rai ds for horses and camels . Otterbein prudently separated Naroll ' s attri bute of 268 revenge and defense, a lumping which obscures a critica l dif- ference between defending oneself when attacked and seeking revenge for a felt wrong . Moreover , not all peoples , "1var - ring " or otherwise , defend themselves when attacked or seek revenge thereafter . The assumpt ion tha t t hey do is just that--an assumption . While theoretically a society would not exist for l ong if it di d not at least defend itself , any deterrent effects of revenge may not be felt immediately . A people may defend themselve s but do not subsequent ly seek re venge, e . g . , the Lepcha, or they apparent l y may do neither i n a certain cultural context, e . g ., the Semai . I mus t admit that Naroll would not identify these 2 peoples as "warring , " even though they have experienced warfare . There i s a second type of "defense " : defending oneself by attacking first , if it is thought that an enemy is pl anning to attack . While this not i on of defense is familiar in modern international relat ions--the "preemptive strike ," "protective reaction"-- it is a lso found i n tribal societies such as t he Caingang and perhaps has much wj_der occurrence than i s i ndicated in the ethnographie s . An alternative would be to code such a ction as revenge , for an anticipated ac t in this case , but thi s would obscure an intere s ting di st i nction . Thus , whi le Naroll has 4 attributes and Otterbein has 6 , I have used 8 here : subjugation of territory and people ; collection of tribute ; l and- - fields , hunt ing territories , fishing territories, pastures ; plunder (includi ng captives 269 for foo d , slaves , hostages , adoption , and marri age ) ; trophies and honors ( includi ng captives for sacr ifice) ; reve11ge ; defense ; a ggressi ve defense (defendi ng oneself by attacking first) . I have handled these at tributes 4 ways : each as a separate variabl e i ntercorrelated with a ll ot he r variables but not other military expectations ; lumped i nto 4 l arger categori e s ; and scaled in 2 dicho tomi es . Both di chotomies are separated i nto hi gh military expectations contrasted t o lo~ mili t ary expectations . In the first scal i ng , i f subju- gation , tribute , l and , or t rophies and honors are pre s ent f or a g iven s ociety, that society is rated as having high mili - tary expectations . In the second scaling , only i f subjuga- tion , tribute , or l and were present i s a society given the high rating . Since in previous s calograms prestige sorted out as a higher expectation than plunder , revenge , and defense , I wanted to see what effect if any i ncludi ng tro- phies and honors would have on patte rns of f requency i n war- fare . My ethnographi c cases of inspiration are the mounted I ndians of North .Ameri ca and pastorali sts of Af r ica , t he Ci rcum- Mediterranean , and East Eurasia . I n co ding these attributes , I counted as present i f i ts presence was specif ically mentioned by t he ethnographer . I f it was no! ment ioned , I counted an expe ctation as absent rather than "no data . " Occas i onally the ethnographers a l s o ment i on the reasons people do not go to war . The c odi ng of defense was generally difficult because so few autho rs 270 specifically ment ioned it. Following the rule , I marked it as absent . I simply do not know where it was truly absent and where the ethnographer assumed that any ninny would know that it would exist . Otterbein (1 970) also had difficulty cedi ng this attribute because of such an assumption . There - fore , I do not consider my quantitative results on defense to be reliable . Otterbein made one useful di stinction that I could not implement satisfactorily . He coded the military expe ctatio ns in order of i mportance--first , second , third . There were too many cases where I co uld not make such a judgment , even if the data on expectations per se were good . 31 . Estimated casualty rate . I modified this variable from one of Otterbe in 1 s, in which he was interested in casualties sustained by a society using its major form of fighting as a measure of efficiency . I am interested in casualty rates regardless of mi litary modes . This has turned out to be a difficult variab l e to code ; i nformation does not exist i n most cases . The dichotomy adds furthe r complica- tions : Otte rbe in ( 1970 : 81 , 146 ) rated as "high II the deaths of over one- third of the combatants and lower losses as "low." There is irony here along with difficulty . Even i n the state wars of the 19th and 20th centurie s , casualty rates of such magnitude that include the woun ed and missing as well as the dead are unusual. During World War II , when losses were 2 to 4 times greater that in Wo rld Wa r I, whi le Germany lost 3 . 35 million of 10 . 20 million mobilized , 271 Japan lost 1 . 51 million of 9 . 9 million mobili zed , a surpri s - i ngl y small overall figure gi ven the low survival rate of J apanese sol di ers i n specifi c battles . In Wc r ld War II, with total losses estimat e d a t 60 million persons , 17 million were military and the remainder were civilians dying most ly i n military a ction but a ls o as a result of war-induced epi- demics (Wright 1965 :1 542-43 ). Loss of one-third of the combat a nts must be rare indeed . Ot t erbein encountered dif- ficulty i n co di ng this variable , als o. Of the 50 societies i n his 1970 sample, he co ded casualty rates for 33 , and of t hose 13 a re high . I would disagree with several of those decisions , applying Otterbein ' s di chotomy strictly . Where the i nformation existed , I code d for the casualty rate of all me mbers of a party to war , maki ng no di s tinction be t ween combatant and non- combatant . Even so , I was able to code casualty rates for only 79 of the 186 societies in the Standard Sampl e . 32 . Command. While Otterbein (1 970 : 23- 28 , 144) des i gned a variable around the "degree of subord i nation " i n a s ociety 's military organi zation , agai n with military efficiency in mind , I am more i nterested i n organi zation to exa ct compli- ance--do persons behave differently under condit ions of i ndi - vidua l autonomy t han they do under conditions of superordi- nate authority? The wordi ng of our variables I think r ef l ects well the difference i n theore t ical orientat ion be tween Otterbei n and me : 272 Wr,.at is the degree of subordi nat i on within the mili ta:cy organization? ... 1 . high- - warriors obey orders gi ven by leaders 2 . low--warriors frequent ly do not obey orders given by l eaders (Otterbein 1970 : 144 ) Who gives commands dur i ng battle? 1. an official who can back up his decisions by force 2 . an i nformal leader who~ people obey because of re spect , but who has no means to f orce warriors to obey 3 . eve ryone is on his own . When I drew up this variable , I negl ected to allow for those cases where a leader has an official position-- appointed war chief, hereditary war chief--and is very aware of the honor and respect of this office but st i ll has no means to force followers to obey. His formality may be ho l - low if the people choose to disregard his direction . 33 . Prestige of a warrior in the community . I have designed this variable to code attitudes with regard to the social status of a warrior or soldier , assuming t hat high prestige will a ct as a stimulus to participate in military activities a nd thus strengthen the mili tari sm of a given society . Tbere i s a certain ambiguity i n the first attri- bute-- "a great deal ; import for every male "--that i s par- tially resolved by the other att itudi nal variables. For instance, Plains Indian warriors accrued enormous prestige through military exploits , carried out with nonauthoritarian leaders . I n modern We stern states , it i s important for every male to do his military service if conscripted and he will be severely punished if he does not , but the personal prestige 273 of soldi ers is relatively l ow . 34 . Cowardice. This variable is a fa i lure , but for an i nterest i ng reason . With very few exce pt ions , even i ncluding professionally trained ethnographers , obs ervers simply do not state what a ction brings down an accusation of cowardice . Cowardice is often ment ioned , but not what i t is seen to be . It must be a uni versal by assumption and not sub j ect to relat ivist s crutiny . 35 . Rewaras for warri ors . This variable augments vari- a ble 33 . I am interested i n t he extent t o which members of t he community publicly reward the military behavior of indi- vidual s . I ask , "Were there special gifts , pra ises , or cere- monies (Lot including ritual purification) for a man who has killed an enemy i n battle or otherwise shown skill in war?" ~he att ributes were yes , u sually or a lways (with the i mpl ica - tion of elaborate r i tual); somet i mes (with the i mplication of perfunctoriness a nd simpli city) ; rarely or never (with the i mplication of minimal, occasi onal , or total absence of such rewa rds) . As in variable 33 I ho ped to collect data to sup- port the argument that rich rewards are an i ncentive t o par- t ic ipate i n militarist activity . The ambiguity of t he state enters he re , a lso . Soldiers in state armj_es may receive recogni tion of personal actions ( ribbons , medals , citations , booty ) , but it i s agai nst the l arger background of the mass of soldiers pe rforming thei r soc i a l duty . The i ndividual so l di er i s only occas ionally singled out . 274 hitual puri f ica tion i s an ent i rely different matt er f ram rewaro_ s . In retrospect , I shoul d have i ncluded a sepa- rate variable for ~he presence or absence of such r ituals , since I have a hunch that the more seri ous l y members of a culture regard the taking of human life as undesirable , the more elaborate wi l l be rituals cleansi ng a person from carry- i ng out the most heinous a ct . Conversely , the l ess seriously a community regards such a transgress ion , the less elaborate will be the rituals . I n the Indo china Wa r , the American mil- i tary and civilia n community omitted such soci al puri fic ation altogether , with mi se r y for the men as a conseque nce . Per- haps such an omi ssion i s cons i s tent with our modern disregard for human life , even within our community . 36 . Expe ctations of violence . I wanted to somehow test Gordo~ Allport ' s notion , discussed i n Chapter I, part 2 , that i f people expect violence to so l ve their problems , they will use it . My operat ionali zation of this vari able breaks all the methodological rule s and subsequent ly i s vulnerable to a tt a ck . The attributes are s imply yes, no , and ambivalent or no evidence , and the coder has used hi s assessment of the material overall. But le t me defend myself a little . Some- one has suggested that I coul d use policj ng and l egal i nsti- tutions as a moasure of expectations , but we only have to lo ok at the United Sta te s today to fi nd objections to such a measure . Wi th a ll our i nstitutions fo r socia l control and legal redres s , our cul tural violence i s legendary . And as 275 other cultural patterns diffus e , so has violence, to Great Br itain for i nstance . A count ry like Lebanon , on the other hand , while it has a system of Roman l aw inherit ed from the French that Americans would find oppressive , exists in a con- dition that we woul d cal l anarchy . Yet the i ncidence of per- sonal and inst i tutional violence i s very l ow . The army prior t o 1972 was a national j oke . The police are i nterested pri- marily in equal sectarian representation on the f or ce . I t hink the variable is enormously important , and eventual ly I hope to bui ld some rigor (but not mortis ) i nto i t . 37 . The value of violence . This variable complements but does not duplicate , or resolve , vari able 36 . The question is , "Is violence/war against non- members of the group ... ( a ) en joyed and cons idered to have high value ; ( b ) considered to be a necessary evil; ( c ) consistent l y avo i ded , denounced , not engaged in ." The 3 attributes are dichotomized by contrast- i ng (a) t o (b) l umped with ( c ). Clwnsy as thi s vari abl e may seem , it was quite easy to code , l argely due to clear obser- vations by the ethnographer as to the val ue of war . The evi- dence may be comments by the observer , texts of poems and songs , or statements by the actors . 38 , 39 . Mi l itary success . I borrowed this variable fr om Otterbe i n . Until going over his material I had not consid- ered including such a variable at a l l , but after the fa ct i ts necessity seems obvious even if only to make tests of Otter- bein t s conclusions possible . I modified Otterbein ' s wording 276 and i ncluded popula tion as we l l as territory as the entities to be · adjudge d as expanding , rernai.ni ng stationary , equa l - i z i ng , or shri nki ng , whi l e retai ning territory as the pri mary meas ure . Otterbe i n dichotomi.zes t he variable only one ·way , as I have discussed i n another s ection , assuming t hat any society that i s not actual ly expandi ng terri torially i s a military failure . I have used the same dichotomy but added a second scaling , assuming that a ny soci ety t hat was not a ctu- ally shri nki ng te rritori a lly was a military success . The second scaling has proven to be of litt le use , s ince i n a fourfo l d table the c cell as we l l as the a cell is loaded up by such a reclassification of attributes , loweri ng the corre - l a t i on coefficients . Thi s concludes my presentation of the basic variables and t he ir variations that I have i ncluded i n this quant i- tative s t udy . An exhaust i ve i nventory of correspondi ng hypotheses i s not necessary , i n my e st i ma ti on , be cause of the stated purpos e of this study . There are several things hap- peni ng here . Avowed l y this study began as a broad unde r - taki ng , us i ng i nductive techni ques to discove r patterns or the absence of patterns among the numerous sociocultur al and warfare variable s - -Harris ' throw-it-against - the - wall- and- se e- i f - i t - st icks technique . Throug110ut t he first chapter I have pre sented and criticize d many hypotheses and theori es of war- f are , s ome of which a re t estable i nduc tive l y . I have done 277 detailed critici sm of other qua nt itat ive warfare studies , emphas i zing t he i r weaknesses but pointing out st rengths as well . Since several of the variables i n this study duplicate or approxi mate tho se used i n the smaller studies , one aspe ct of my work i s to retest selected variables on a l a r ger , more rigorous sample . I have attempted to explore any relat ion- s hi p between economic patterns and warfare , a dry hole in OtterbeiL ' s 1970 work . I have looked at community organiza- tion and populat i on density , basic demographic vari ables . I have tried to isol ate basic so ciocultural variables a s the background agai nst which warfare can be underst ood , holding as a possibility that there may be some configuration in which warfare is the i ndependent , no t the dependent , vari - able . I have tried to do more on a bigger canvas than anyone else has doue on warfare in anthropo logy . There are , moreover , 2 addit i onal hypotheses that I am clearly te s ting: the existence of significant geographi- ca l differences obvia ting a unive r sal explanation, a nd the associat i on of attitudes and values held about warfare by the actors themselves with the other variabl es . I will admit again that the attitude variables are wanting in operat ional - i zat ion, and that I shoul d have included more and better variables . At the same time , I will defend t heir inclusion as going i n a better di re ction than more traditional , psycho- logical var i able s , and that a pre limi nary test may hint at future i mprovements . I may be critici zed that the attitude 278 variables ara loaded--that I am bound to get what I 1-rant i nductive l y be cause of t he co~ing which i s biased i n favor of my a rgument . There may be cons ide rable truth to such a criticism on the basis of the overt crudity of the variables. The broadness of the variables is partly due to 1 other facto r . Readi ng and coding warfare cat a on t he so cieties in the Standard Sampl e was a l ready well along before I began to develop an i nterest i n and argument fo r t he significance of beliefs and values in war . I had already c hanged the vari- ables twice and simply could not expe nd the effort re qui r ed to begin again and do elaborate att itudi na l resear ch , even if the i r_fo rmat ion exj_ s ted i n the literature , which it usually does not . In this chapter I have presented the justification, precedents , methodology , and substance of the original re search i n this di ssertat i on . In the third and last chap- te r, I shall report my f indings . 279 CHAPTER III FINDINGS : DISCOVERIES OF THE ODYSSEY I n a whims ical moment , the computer program for this research was christened ODYSSEY . Afte r a time the name seemed more and more appropriate--better than AGONY , or BLOOD , or GORE . So, if the readers will cont i nue to be tol- erant of my poet ic i ndulgences , they may relieve the i nevit - able tedium of statist ical dis cussion . I will first present tne quantitative results of the ' round the world i nspection of war , and then look closely at each of the 6 geographical regions . The model I am us i ng is an art icle by Driver and Schuessler (1 967 ) i n which they intercorrelate and factor analyze 30 sociocultural variables-- various subsistence activities , settlement patterns , forms of marital residence and descent , kin terms , political iritegrat ion and succes- sion--coded by Murdock i n his World Ethnographic Sample (1 957) for 565 ethnic units taken a ll together and then wi thin each geographic region . A View of the World Before looking more closely at the s i gnificant asso ci- ations without regard to relat ive geographical distri bution , I would like to discuss the results of geography as a vari- able . The C values for these i ntercorrelations are i n 280 Table 4 . In the interest of avoidi ng repet i t i ous ~ording , I shall i nventory the var i ables that are s i gnifi cant at the . 05 level or hi gher wi th each geographi cal regi on and their pre- dic t i ve d i re ct i on . When I state t hat a var iabl e i s present or absent at a ce r tai n s i gni f i cance , thi s does not mean t hat it i s a l ways pr esent or a l ways absent f or every s ociety . It means t hat cert a in a tt r i butes of a vari a ble are a bsent often enough t o mat hematica lly domi nate t he number of time s they are present , tha t t he a bsences occur more often t ha n i f by chance . The mat hemat i cal statements ~re simpl y sunllllari es . The l ambda value s are generall y lower i n t he worl d- wi de i ntercorrelat i ons t han i n t he int e r nal regi onal one s . 1 . Afri ca . Polygyny i s present ( . 001) and the predict ive di r ec- t i on of l ambda i s the presence of po l ygyny from knowl edge of "Afri ca" 21 pe r cent of the time but no a bili t y to pr edict Africa from po l ygyny . Compact but i rnpermanent se t t l ement s , neighborhoods of di spersed family homesteads , and separat ed hamlets are present ( . 05 ) and patri lineal descent i s present ( . 05) , but ne i ther attri bute possesses a predict i ve dire c- t i on . Unilineality i s pr esent (. 05) . I n the domi nant sub- sistence activi ty , divi sion of l abor i n whi ch men do all or most of the work i s absent ( . 02 ) but i t i s not predictable . No other basic s oc iocultural vari ables asso ciate s i gni fi - cant l y with Africa . 281 TABLE 4. COMPARISON OF C COEFFICIENTS: GEOGRAPHIC REGION AS A VARIABLE Variable Africa Cir-Med Eurasia Pacific N Amer S Amer 2 ( 0 335)* .042 (.239)* .071 .052 .020 3 .116 (.235)* .005 .007 .125 (.198) 4 .032 .100 .089 .006 .046 .006 5 .132 .111 .134 .o84 .124 ( .154) 6 .066 .076 .012 (.221)* (.311)* .039 7 (.166) .028 .103 .094 (.198)* .085 8 .074 .094 .094 .137 ( • 145) .106 9 .029 ( .259)* .024 .086 .070 .101 10 .124 .083 .051 .068 .o4o ( .164) 11 .090 ( .167) .026 .142 .026 .062 12 .072 .031 .088 .102 .055 .127 13 .o48 ( .350)* .117 .056 (.226)* (.214)* 14 .131 (.296)* .113 .031 (.225)* ( .328)* 15 .107 ( .249)* .062 .103 ( .366)* .114 16 (.201) .176 .095 .136 ( .318)* .098 17 ( .144) .051 .140 .010 ( .177) ( .165) 18 ( .179) ( .182) .024 ( .152) (.204)* .050 19 .068 (.245)* .045 .034 ( .155) (.223)* 20 .090 .007 .040 .025 .067 .009 21 .074 .o44 .001 (.199)* (.312)* .123 22 .132 ( .155) .076 ( .178) .089 .050 23 .090 .071 .118 .099 .005 .051 24 .078 (.209)* .058 ( .176) .071 .089 25 .018 .125 .069 .060 (.227)* .027 26 .015 .130 .080 (.247)* .148 .148 27 .053 .027 .074 .075 .057 (.225)* 28 0 .067 0 (.298)* .081 .205 29 .003 ( .186) .008 .003 .022 (.194)* 30 .021 (.285)* .091 .030 .107 (.217)* 31 ( .216) (.290)* .091 .072 .047 .056 32 .042 .118 ( .185) .025 ( .294)* .010 33 (.200)* .033 (.212)* .001 .079 .094 34 .165 .152 .238 .075 .168 35 .133 .065 ( .206) ( .189) .077 .035 36 .076 ( .178) ( .187) .091 .002 .010 37 .065 (.201) (.168) .065 .068 .080 38 .123 .113 .107 .106 .093 .o84 39 .057 .013 ( .158) ( .185) .001 .056 40 .056 (.YJ7)* .067 .040 (.208)* .061 41 .062 (.204) * .076 .055 .104 .020 42 .042 ( • 116) (.574)* .056 .o44 (.236)* 43 ( .162) .012 (.255)* .027 .129 .010 44 .o48 .134 ( .1 85) .050 ( .282)* .045 45 .1 20 ( .170) (.226)* .087 ( .314)* • 106 46 .005 .113 .041 .014 .106 ( .182) 47 .056 .061 .137 .050 .oo4 .092 ( .E = .05) ( .E = .01 ) -i<· 282 Among the warfare variables , the presence of external war-attacking is non- signifi cant(< .1 0 > . 05) . High casual- ties a re s i gni f icant ly absent ( . 05 ) but the absence i s not predictable . High prest i ge f or warriors i s present ( . 01) and i s predictable from "Africa " 10 . 8 per cent of the time . Plunder as a military expecta tion i s pre s ent (.05) but not predictably . 2. Circu.111- Medi terranean . Thi s ge ographica l regio n and North America have the l argest numbe rs of significant associat ions with the other variabl es . Patril ocality is present (.01) but not predict - a bly . A mean community size of over 400 persons is present ( . 001) , predictable from the Circum-Mediterranean (. 077). Two or more levels of j urisdictional hi erarchy beyond the l ocal community are present ( . 001) and knowl edge of this is predictabl e fr om knowl edge of t he Circum-Medi terranean ( . 196 ) . Food- producers are present ( . 001 ) but not predict- ably . Thi s i s odd be cause food- production i s the only sub- sistence mode i n the region . Men doing most or ail of the work in the domi nant subsistence a ctivi ty i s present ( . 02) , and it i s predictabl e from knowledge of the region (.11 2). Class stratification is pres8nt ( . 001) and predictable i n the Circum- Medi terranean ( . 102) . This region has numerous s i gnificant warfare a ssoci- ations . While the l ambdas remain weak , they are stronger 283 here than in associations with other regions. External war- attacking is present (.05) but not predictable. Mobilization organized through age-sets, military societies, or armies is present (.01) and predictable (.095) . High military expec- tations defined by subjugation, tribute, or land are present (.001) and pre dictable (.207). High casualties are present (.01) and predictable (.194). Violence is expected to solve pro~lems (.05 ) but its presence is not predj_ctable. Violence is highly valued (.02) and predictably so (.147). The fol- lowing mib.tary expectations are present : subjugation (.001; A= .027) , trib~te (.01; not predictable), and land (.05 ; not predictable ). Revenge as a military expectation is absent (.05) and the absence predictable (.047). There is, however, no associatioD wi. th aEy £.:rm of military success. 3. Eastern Eurasia. This region produced sigI1ificant asso(!iations that a:::.'e strikingly different from those of the Circu.rn-Ilfodi terranea:n. . There is only 1 sociocultural variable : - polygyny i s absent (.001) but not predictably so . Among the warfare variables , authoritative command is present (.05) and predictable ( . 119). High prestige for warriors and soldiers is absent (. y1 ) but not pre djctably. Elaborate public rewards for warriors are absect ( . 05) pre- dictabl.y (. 095) . Violence is not expe e;ted to scl.-.,re pro blerns (.05), and i ts absence is predictable (.129). War is 284 regarded as a necessary evil or denounced and avoided ( . 05 ), pre dictabl y ( . 093) . Military succe s s defined as expanding, unchanging , or equa lizing i s absent ( . 05 ) but not predictably. However, there is no significant associat ion between military succes s defined as expansion alone and Eastern Eurasia . The followi ng military expectations are absent : plunder ( . 001 ; A = . 147), trophies and honors ( . 02; A= 0) , and revenge C ( . 01 ; Ac= . 094 ). 4 . Insula r Pacific . Nomadic , semi-nomadic, and. semi- s edentary settlements are absent in this region ( . 01) but not predictably . Women do as much work as t he men , or more, i n the domi nant sub- sistence activity ( . 05) , but not predictably . Among the warfare vari able s , internal war is present ( . 01) and predictably ( .132) . External war - attacki ng is a bsent ( . 02 ), with a small predictive value ( . 088 ) . Military organi zation i n age - sets , military societies, and standi ng armies is absent (.02 ) but not predictably . War begun by a greement or announcement i s present ( . 01); a peace ceremony is present ( . 01) ; rewards for warriors are pr e sent ( .05) ; mi litary success de f ined as expanding , unchanging, or equal- izing i s present (.02). None ·of these 4 s i gnificant corre- l ations i s . predictable i n either direction . 285 5. North Ameri ca . Nomadic , semi - nomadic , and seff1i- sedentary settlements are present (. 001) and predictable from knowledge of "North America" ( .1 77 ). Homesteads and scattered haml ets are absent (.01) but not predictably . Compact and complex settlements are a l so absent (.05) and the absence is weakl y predictable (. 024 ). Any levels of jurisdictionary hierarchy beyond t h e l ocal community are significantly absent (.001) with a re la- tively strong predictive value (.232 ) . Foo d-collectors are present ( . 001) and predictably ( . 143). Matrilineal descent , as oppo sed to only patrilineal des cent , i s s i gni f icant l y present ( . 001) and predictably ( .1 47 ). Bilatera l descent, as opposed to unilineal descent , is present (.02) and predict- a bly ( .1 25) . Men do all or most of the work in the dorr,i nant subsistence activity ( . 01 ), predictably if they are North American (.1 38 ). Class stratification i s absent (.05) but not predi ctably . This pattern of s ignificant presence s and absences of specific variables i s in strong contrast to the configurat i ons i n the Ci rcum-Mediterranean region . The warfare variables show another distinctive pat tern, actually confirmation of the common sense conclus ions one develops from s i mply reading the ethnographie s . The numbers show that internal war is rare enough to be significantly absent i n North America ( . 001), and predictably (. 118) . Off icia l decis i on- making for g oing to war i s absent (. 01 ) , 286 whi le beginning a war by surprise at t a ck i s present ( . 05 ) , but the associat ion of neither vari able is predictable. Warriors fo llow an i nforma l l eader or are on their own in battle (.001) and this rela tionship is predictable (.09) . The following milita ry expectat ions have a significant ass o- ciation with North America . Trophies and honors are present (.001) predictably ( .1 00 ). Revenge i s present (.001). Sub- jugation of territory and people is ab sent (< . 01). The last 2 as s ociations are not predictable . 6 . South America . lfatrilocal re s i dence is present(< . 02) when opposed only to patrilocal residence i n the correlation with this geographical region, but there i s no predictive direction . Community endogamy is pre sent (< . 05), not predictably . The average size of a typical community of between 0 and 99 per- sons is pre sent ( . 05) with a very weak pre dictability value (.036) . There are usually no levels of jurisdictional hiera.rchy beyond the local community (.001) and thi s is pre- dictable 23 per cent of the time . Unilinea l descent is absent ( < . 05) predictably ( .1 03 ) . Class stratification i s also absent. Amo ng the associat ions between South America and t he warfare vari ables , conclus i on by negotia tion is ab s ent (< . 01) wi th a small predictive value (.0469) . Low military expectations--indexed by the pres ence of defense , revenge , 287 plunder--are present ( . 01), an d when trophies and honors are also included as a low military expectation , prediction i s . 099 . The specific military expectat i ons of land and defense are s i gnificantly absent ( < .01 and< . 02 respectively) but only defense generates predict i ve direction ( . 105) . Thus, by using geographical region as a variable i n t he worl d- wide intercorre l ations , disparate inventorie s of variable incidences exist among the regions . Many of them are familiar, for i nstance polygyny and women in subsis t ence work in Africa; foo d-product ion , class stratification, patri- l ocality , pat ri- dominant labor , and the state in the Circum- Medite rranean ; small , relatively transitory communities , the ab s ence of complex political organization , food- collectors , and bilaterality i n North America . As for the warfare vari ables , the configuration in the Circum- Mediterranean i s i mpressi ve : external war- attacking takes pl ace with formal military organi zat ion , going afte r high mi l itary expectations ( subjugation of people and terri- tory , the collection of tri bute , and land ), sustaining hi gh cas ual t i es , and i s seen as a solution .to problems and highly valued . In Eastern Euras i a , however, while there is authori- t ative command i n war , there are no elaborate rewards or high prestige for warriors , and violence and war are not adjudged as solutions to problems but seen as necessary evils or avoided . I n yet another variation, in the I nsul ar Pacific internal wa r is the mode , carried out by agreement by the 288 parties , conc l uded wi th a peace ceremony and rewards for the warri ors . I n this region cultures are militari ly successful i n the sense that they are not lo sing terri tory . Probably the island environments of so many of the cultural units in this region a cc ount for this warfare pattern . Worl d Confi gurat ions Looking next a t t he worl d sample as a whole , every variabl e has been i nt ercorre l ated wi th every other variable but not with it s own vari ation , e . g . , the 2 scalings of the marital residence attributes are not correl ated wi th each other . The Pearson ' s C coefficients for each compari son are arranged i n Table 5. Those values s i gnificant at the 5 per cent level or higher are i n parentheses . Tho se significant at the 1 per cent l evel or higher are additionally marked by an asterisk . Out of a total of 987 i ntercorrelat ions , 308 or 31 . 5 per cent are s i gni f icant at the 5 per cent l eve l or hi gher . Two hundred s ix or 20.8 per cent are s i gnificant a t t he 1 per cent level or higher . I n contrast , out of Driver and Schuessl er 1 s 435 C coeffici ents , a mere 7 . 8 per cent are s i gnificant at t he 5 per cent level or higher, a surprisingly .low figure . Furthermore , No one singl e variable i s consistent1y correlated with all of the rest , or even with a small subset of them . Except to hint at possible clusters of trai ts , these ca t - egori es would have to be combined i nto broader groupings TABLE 5. .£ COEFFICIENTS FOR THE WORID: EACH VARIABLE WITH EVERY OTHER VARIABLE Variable 2 3 4 5 6 7 8 9 10 11 12 13 14 15 2 3 (.171) 4 ( .143) 5 .110 ( .259)* .111 6 .030 .041 .009 .019 7 .092 .033 .083 .011 8 .103 .013 .059 .027 9 .012 .o6o .037 (.179) (.259)* .o4o (.217)* 10 .039 .010 .087 .106 ( .317)* .063 (.256)* 11 .025 .065 .047 .065 .o64 .022 .042 12 ( .161) .053 (.248)* .015 (.192)* .o44 ( .149) .104 (.234)* .130 13 .014 .107 .047 .125 ( .304)* .130 (.380)* (.322)* (.277)* .037 14 .118 (.172) (.150) .091 ( .287)* ( .167) (.390)* (.320)* (.386)* .086 15 .086 .125 .050 .025 (.522)* (.265)* (.336)* (.334)* (.343)* .016 (.203)* (.334)* (.425)* 16 ( .297)* ( .6,65)* ( .203) (.270)* .077 .oo4 .071 .175 .136 .048 .020 .140 ( .245) .180 17 ( .165) .024 ( .269)* (.237)* (.148) .109 .053 .042 ( .163) .115 .271 .055 ( .238)* ~ .278t 18 ( .151) .119 .038 .071 .105 .o64 .049 .083 .066 .014 .103 .032 .127 .222 * 19 .086 .113 .022 .063 ( .318)* .101 (.377)* (.257)* (.288)* .o46 .037 (.493)* (.467)* (.305)* 20 .103 .005 .o46 .032 .034 .033 .059 .071 .147 (.197) .098 .088 ( .154) .070 21 · ( .157) .069 .114 .036 .127 .110 .034 .085 .034 .047 .114 .001 .107 .117 22 (.220)* .019 .029 .o46 .047 .o84 .111 .138 ( .192) .058 ( .157) (.225)* (.266)* ( .154) 23 .081 .042 .027 .049 .080 .019 .062 .115 .137 .025 .078 .038 (.181) .136 24 .026 .010 .036 .034 (.276)* .149 (.365)* (.329)* (.232)* .103 .oi19 (.423)* (.291)* (.286)* 25 .111 (.180) .018 .095 ( .264)* .034 (.267)* (.185) (.246)* .040 .055 (.226)* (.265)* (.251)* 26 .039 (.176 .024 .022 .101 .070 .oi11 .075 .139 .058 .031 .157 ( .179) .116 27 .143 .009 .011 .027 .044 .126 .o6o .126 .142 .009 .050 .059 .014 .149 28 .083 .007 .037 .049 .061 .128 .055 .008 .o4o .042 .010 ( .263) .043 .016 29 .1 ·11 .018 0 .061 (.205)* .110 (.275)* (.232)* (.270)* .035 .035 (.249)* (.400)* (.253)* 30 .014 .069 .107 .003 (.188) (. ·175) (.307)* (.247)* (.228)* .021 .013 (.328)* (.377)* (.228)* N ( .E == .05 ) ( .E = .01 )* O'.l \.0 TABLE 5,--Continued Variable 2 3 4 5 6 7 8 9 10 11 12 13 14 15 31 .030 .026 .070 .092 .128 (.273)* (.331)* .204 .069 .131 .076 (.322)* (.239) .139 32 (.222)* ,122 .097 .145 (.387)* .043 (,396)* (.258)* (.280)* .011 .005 (.471)* (.411)* (,368)* 33 (.259)* .128 .008 .o4o .056 .018 .069 .oo4 .127 .114 .026 .061 (.204)* .151 34 .094 .102 .030 ( .290) .033 .001 ~030 .077 .o6o .112 .157 .025 .172 .232 35 ( ,243)* .035 .169 ,083 .126 .070 .062 .130 ( •2 14) .072 .061 .066 (.251)* (.262)* 36 (.266)* .062 ,088 •1 58 .132 .056 ( • 171) ( .280) .129 ,070 • 163 (.219)* (.210) .047 37 (. 305)* .114 .115 .089 .114 .082 ( .172) .037 .148 • 104 • 142 (.244)* {.;()7)* (.164) 38 ( , 173) .009 .136 .052 ( .178) .0?7 (.190)* .116 ( .146) .029 .065 (.284)* (.171) .129) 39 .011 .o64 .058 .075 .098 .007 .099 .002 .045 .o44 .051 .031 .009 .077 40 .109 .094 ( .165) .085 ( .268)* .088 (.318)* ( .298)* ( .237)* .062 (. 158) ( .489)* ( -333)* (.291)* 41 .056 .077 ,004 .068 .081 .078 .138 .137 .024 .105 .042 ( ,350)* ( .222)* , 121 42 ( • 172) .110 .oo4 .033 .015 ( • 171 ) .126 .128 (.172) .042 .078 .108 (.278)* .106 43 (.291)* .010 .009 .071 .093 .089 .016 .o84 .020 .060 · .045 ,039 .103 ( .149) 44 (.231)* .045 .034 .092 .o64 .011 .052 .037 .146 (. 168) .101 .051 .126 .071 45 .097 • 108 (.169) .038 ( .204)* .088 ( .259)* .100 .083 .016 .051 (.253)* (.276)* (.271)* 46 .043 .024 • 142 .o6o .013 • 181 (.159) ( .147) .056 .086 .067 .069 ( .150) .110 47 .023 ,075 (,234)* .022 ,060 .036 .086 (,150) .055 (.189) (.289)* ,045 .101 .016 f\) \.0 0 TABLE 5---Continued Variable 16 17 . 18 19 20 21 22 23 24 25 26 27 28 29 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 .110 ( .172) 19 .072 .012 .016 20 (.217) (.233)* .033 (.210)* 21 .054 ( .259)* .132 .077 .014 22 .077 .123 .002 (.222)* .099 (.168) 23 .068 .116 .o84 .045 .094 .131 (.447)* 24 .074 .011 .o64 (.395)* • 138 .083 (.174) .034 25 • 100 .090 .088 (.342)* (.308)* .091 (.247)* .079 (.230) 26 .005 .017 .019 (.222)* (.181) .053 • 140 .114 .o84 .o44 27 .045 .061 .088 .015 0 .057 0 .128 .057 .033 .o6o 28 .221 .028 ( .277)* .16o .069 (.231) .028 .060 ( .298)* .087 .103 (.290)* 29 .065 .008 .091 (.28o)* .oo4 .o44 (.327)* (.220)* (.234)* (.197)* .091 .089 .042 30 .111 .006 .081 (.335)* .075 .022 (.212)* .100 (.281)* (.262)* .052 .009 .069 I\) \.D TABLE 5--Continued Variable 16 17 18 19 20 21 22 23 24 25 26 27 28 29 31 .138 .153 .072 l ,239) .101 .035 l .276) .214 l .230) .188 .101 .095 .141 ( .255) 32 .129 .111 .053 (.449)* (.204) .022 ( .187) .007 (.355)* (.403)* ( .119) .013 .160 (.287)* 33 .189 .098 .078 .118 .067 ( .196) (.292)* (.229)* .032 .057 .059 ( .175) .032 ( .293)* 34 .091 .155 ( .274) .107 .019 .092 .088 .058 .031 .185 .224 .032 .086 .152 35 .085 ( .249)* .116 .086 .076 (,288)* (.449)* (,369)* .126 .172 .022 .020 ( .255) (.309)* 36 .081 .106 .029 (.246)* .o46 (.261)* (.414)* (,212) .085 •0 79 .011 . .162 .163 ( ,333)* 37 .182 ;028 .088 (.277)* .103 ( .173) ( .413)* ( .204) .100 .024 .111 (.250)* .086 (.282)* 38 .085 ,041 .007 (.201)* .049 .098 (,373)* .091 (.178)* (,238)* .054 .085 .075 (.336)* 39 .095 .088 .033 ,011 .018 .030 .039 (,239)* .019 .006 .100 .081 .020 ( .154) 4o .057 .101 .138 ( ,395).* .o4o .056 .086 .049 (,436)* ( .224)* (.172) .023 .181 41 ,142 .003 .055 ( .301 )* .145 .005 (,201)* .079 (.223)* (.179) .134 .117 .087 ~ ..237849 t* 42 .178 .167 .061 ,148 .041 .045 ( .233)* (, 18o) .001 .078 .067 .020 .075 (.456)* 43 .021 .118 .021 .059 ( .198) ( .162) (,353)* (.264)* .107 .070 .079 ( .176) .163 44 .062 .059 .057 .oo4 .072 .031 (.249)* (.226)* .025 .022 .027 ~ .1 .003 .028 .4 5571~ * 45 ( .234) .120 .083 (.240)* .070 ( .167) .003 .095 (,325)* (.264)* .033 .001 .016 .097 46 · .024 .144 .005 ,110 .043 .070 .065 .143 ( .224)* .091 .050 .038 .126 .080 47 .047 .o84 .055 .024 .059 .073 .100 .095 .121 .057 .042 .129 0 .034 I\) \.0 I\) T~ 5--Continucd Variable 30 31 32 33 34 35 37 :I) 31 (.260)* 32 (.?65)* (.249) 33 .120 .070 .010 34 .220 .284 .022 .091 35 .092 .105 .132 (.469)* .101 36 .154 ( .285) .149 ( • 1i37)* .114 ( .331 )* 37 .076 • 117 .106 (.449)* .153 (.354)* (.556)* 38 (.322)* .177 (.194) (.283)* .218 .003 (.302)* (.250)* :I) ( .149) .072 .066 ( .193) .110 .041 (. 211) .108 4o (.483)* (.364)* (.428)* .002 .077 .079 (. 177) .126 (.294)* (.150) 41 (.348)* .169 (.253)* .096 .019 • 155 (. 252)* (.169) (.311)* .106 42 ( .575)* .061 .oo4 ( .161) .118 ,091 .077 .o84 ( .231 )* .031 43 .045 .110 .083 (,321)* .116 (.286)* (.204) (.354)* (.243)* .068 44 .115 .038 .030 (.304)* .178 (.323)* (.243)* (.228)* (.211)* .101 45 (. 167) (.311)* (.373)* .054 .081 .117 .081 .009 .073 .107 46 .130 (,291)* .024 .079 .192 .o44 .040 • 142 .118 (. 161) 47 .049 .128 .078 .031 .018 .087 .130 .069 .098 .019 I\) I.Cl '->l 294 if rela tions amo ng tra i ts are to be detecte d (Driver and Schuessler 1967 : 333) . Subs equently , the aut hors relied upon the purely i nductive sta ti s tica l t echnique of fac t or analysi s to float out clus- ters of variables . I do not seem to have that probl em here . By inspection of the C val ue s , the fo llowing "factors " or c cncentrations of significant assoc i at i ons ca ll for atten- t i on : Polygyny ; permanent or complex settl ements ; patricen- te r ed organization; centralized political organization and stra tification ; external war- attacking ; decis ion- making, mobilization , and command in time of war ; attitudes and val- ues with regard to violence and war ; military success ; high Bilitary expectations and the speci f ic expectations of sub jugation , plunder , and revenge . Looking first at polygyny and i ts association with v;arf a re , if any , one finds the presence of polygyny to asso- cia te significantly with the presence of continual or fre - que nt i nternal war at t he . 05 l eve l, and i nternal war is e xplained by polygyny 9 per cent of the time . Otterbein (1 968 ) has u sed polygyny as a n indicator for the existence of fraterna l interest groups and also found a po s itive corre- l ation with internal war at the .05 level. But the l arger s ample r eveals something tha t Otterbein ' s smaller one cannot . Otterbein ' s table (1968 : 281 ) looks like this : 295 I nternal War Cont i nual Infrequent or Frequent Polygyny present 1 5 3 Polygyny absent 1 3 1 1 42 cp = . 31 x2 = 3 .93 p = . 05 Compare thi s t o the resul ts based on the Standard Sample shown i n Table 6 . Note the geographical distribution of the societies in the 4 cells summarized as fo llows . a b C d Africa 1 2 8 2 4 Circum-Me diterranean 7 3 3 1 2 East Eurasia 1 2 1 3 1 5 I nsular Pacific 4 4 1 5 5 North America 2 8 2 21 South America 8 2 7 8 Looking even more closely at the identities of the societies , one sees that those with both polygyny and frequent · i nternal war are African and South American , and Circum- Medi terranean societies that are either located on the African continent or are Islamic or both . There are a l arger number of societ.ies with frequent internal war but without polygyny , and t hey are concentrated i n Eastern Eurasia and the I nsular Pacifi c. The environmental circu.mstances of 1var i n the Pa cific have already been touched upon . Otterbein considers polygyny only as a basi s for mili- t ary organization . Diva l e entertains polygyny as a catalys t 296 TABLE 6. INTERCOlliiliLATION OF POLYGYNY AND INTERNAL WAR Continual/Frequent Internal War Infrequent Internal War Ila Teda · Thonga Bellacoola Nyakyusa Gheg Lozi Yokuts Banen Kazak Mbundu Hidatsa Tiv Tiwi Kikuyu Pawnee Ibo Kapauku Ganda Natchez ~ Q.) Ashanti Ajie Nkundo Comanche 0) Q.) Mende Maori Fon Chiricahua Apache pf...t Tallensi Aleut Bambara Papago Azande Yurok Fulani Saramacca t Nuba Goajiro Hebrews Siriano >, Shilluk Yanomamo Rwala rl Masai 0 Jivaro Gond p. Wolof Tupinamba Chukchee Hausa Botocudo Aranda Kanuri Shavante Kwoma Konso Aweikoma Tikopia Somali Mapuche - 34 Marquesans 27 Nama Palauans Kung Yukaghir Inca Kongo Ifugao Luguru Javanese Trumai Tuareg Mbuti Badjau Timbira Lapps Ingassana Alorese Lengua Kurd Fur Manus Basseri Amhara Gilbertese Kpalka Nubians Ingalik Lola Egyptians Cp. Eskimo Garo Babylonians Montagnais Serna Naga Turks Micmac Burmese Romans Saulteaux Vietnamese Basques Slave ~ Siamese Irish Kaska Q.) 0) Andamanese Lapps Twana .0 < Negri Sembilan Russians E. Pomo Chinese Annenians Paiute t Ainu Punjabi Klamath >, Gilyak Toda Kutenai rl 0 Balinese Santal Omaha p. Iban U. Pradesh Huron Orokaiva Lepcha Creek Kimam Palaung Zuni Lesu Khmer Havasupai Trobrianders Semai Huichol Siuai Nicobarese Aztec Pentecost Vedda Popoluca Mbau Fijians Tanala Miskito Samoans Manchu Warrau Majuro Koreans Carib Yapese 42 Japanese Mundurucu 65 Q. = • 322 /, = • 159 X2 = 4. 262 .E < • 05 ~ ~ = • 09 C = .157 297 i n his warfa re syndrome and raiding for women to offset a wife shortage as t he cause of pri mi t i ve warfare . One bit of i nformation from the ODYSSEY i s that of the 34 po l ygyni st societies that carry on i nternal war 28 , or 82 pe r cent , go t o war for plunde r . Taki ng captives for slaves , adoption , or wives has been defined as one type of "plunder . " I have not t abul ated the number of cases where captured women are we d , a l though I have t r ie d to systemat ical ly colle c t such data . However , 26 ( 62 per cent ) of the 42 non-po l ygynis t societ ie s that wage freque nt i nterna l war also fight for plunder ; t hus , the particular information on women i s necessary t o test this out . The r e i s another, broader perspe ct i ve from which the a ss oc i at ions with polygyny may be viewed . The presence of polygyny i s s i gnificant ly associated with the absence of pat ri- domi nant l abor , the absence of authoritative command , a ~d the presence of patri l ocality , patrilineality , 3 or 4 leve l s of loca l jurisdictional hie rar chy , frequent externa l war- a ttacking , high prestige for warriors , elaborate rewards ' for warriors , violence as a solut io n to problems , violence and war as highl y valued , military success a s territor i a l expans ion , and the spe cific military expectations of pl under and. t rophi es and honors . Te ntat i ve l y I will ca ll thi s an androcentri c configuration , s i nce the common "factor " among t hese vari ables seems to be i ndi v i dual male a chievement and loca l sociopolitical domi nance . Seventy per cent of those 298 societies with polygyny and fre quent internal war also engage in freque nt offensive external war , although the correlat ion between polygyny and external war- attacki ng is the stronge r one at less than the .01 l eve l but without predictive power. Out side of the 70 per cent of shared cases for common presen- ce s , the distribution and i dentity of societies i n the rema ining cells of the intercorrela tion between po l ygyny and external war- attacking are quite different (Table 7). The geographical representation of the 4 cells i s : a b C d Africa 1 6 2 2 4 Circum-Mediterranean 8 1 12 5 Eurasia 2 1 1 5 1 5 Insul ar Pacifi c 2 6 5 7 North America 7 3 9 1 3 South America 8 2 12 9 Africa has an even stronger inf luence on this i ntercorrela- tion , South America t s remains the same (although 1 of the 8 contributing societies i s different) , and North America makes itself felt . Seven of the African s ocieties with polygyny but not frequent interna l war do engage i n frequent offensive external war . While only 3 Circum-Me diterranean societies without polygyny carry on internal war, they and 9 others do attack out s i ders . The foll owi ng series of 2 by 2 tables (Table 8 a - 1) summarizes the significant associat ions of polygyny with the other variables i n the configuration . I have g iven counts 299 TABLE 7. INTERCORRELATION OF POLYGYNY AND EXTERNAL WAR-ATI'ACKING _ Continual/Frequent External War Infrequent External War Thonga Teda Mapuche Nyakyusa Lozi Hebrews Tallensi Mbundu Rwala Gheg Ila Kazak Gond Kikuyu Chukchee Tiwi Ganda Ajie Aranda Banen Marquesans Kapauku Tiv Aleut Kwoma Fon Hidatsa Tikopia Ashanti Pawnee Maori Mende Natchez Bellacoola t . Bambara Comanche Yurok >, Azande Chiricahua Apache Yokuts rl 0 Nuba Papago Goajiro p.. Shilluk Yanomamo Siriano Masai Saramacca Wolof Jivaro Fulani Tupinamba Hausa Botocudo Kanuri Shavante Somali Aweikoma 43 15 Nama Japanese Timbira Kung Alorese Trumai Ingassana Ainu Lengua Kongo Manus Cayua Fur Yukaghir Abipon Luguru Trobriand Yahgan Amhara Javanese Tehuelche Mbuti Marshalls Bisharin Iban Songhai Yapese Tuareg Orokaiva Nubians Palauans Riffians Pentecost Turks Ingalik Anc. Egypt Trukese Basques Cp. Eskimo Babylonians Micmac Lapps Montagnais ~ Romans Eyak Toda Saulteaux Q) 0) Irish Haida Santal Slave ~ Russians Klamath Lepcha Kaska Abkhaz Gros Ventre Bunnese Twana t Kurd Huron Palaung E. Pomo >, Yurak Creek Rhade Paiute ,-I 0 Basseri Zuni Semai Kutenai p.. Punjabi Aztec Nicobar Omaha u. Pradesh Maya Andaman Havasupai Burusho Miskito Vedda Huichol Khalka Bribri Tanala Cuna Garo Callinago Negri Sem: Hai ti ans Serna Naga Mundurucu Chinese Warrau Vietnamese Inca Koreans Carib Khmer Aymara Gilyak Cubeo Siamese Nambicuara 55 Badjau Cayapa 53 Q = .468 ,5 = 2 .225 x = 8.407 C = .220 p < .01 TABLE 8 • . INTERCORRELATIONS OF 'NELVE OTHER VARIABLES WITH POLYGYNY ( a) (b) Postmarital Residence Descent Patrilocal Matrilocal Patrilineal Matrilineal 17 Africa 1 A:frica 16 Africa 11 Circum-Mediterranean 9 Circum-Mediterranean 3 East Eurasia ~ 2 East Eurasia s;:: 8 Insular Pacific Q) 4 Insular Pacific 1 Insular Pacific 0) 6 North America 3 North America Ill 1 North America M p.. 2 South America 5 South America 3 South America 4 South America 47 9 35 5 3 Africa 2 Africa 2 Africa 2 Africa 10 Circum-Mediterranean 5 Circum-Mediterranean 2 Circum-Mediterranean 20 East Eurasia 8 East Eurasia 16 East Eurasia 5 East Eurasia 11 Insular Pacific 7 Insular Pacific 5 Insular Pacific 6 Insular Pacific 12 North America 8 North America 2 North America 6 North America n2 South America 7 South America 4 South America 3 South America 68 32 34 24 A r • .025 \..N 0 0 TABLE 8--Continued ( c) ( d) Levels of Local Jurisdiction Division of Labor in Dominant Subsistence Activity 0-2 Levels 3-4 Levels Males Alone Or More Females Equal, Alone, More 5 Africa 15 Africa 3 Africa 17 Africa 11 Circum-Mediterraneat 4 Circum-Mediterranean 5 Circum-Mediterranean 3 East Eurasia 2 East Eurasia 1 East Eurasia 1 Insular Pacific 7 Insular Pacific ~ 2 Insular Pacific 6 Insular Pacific Q) 10 North America en 5 North America Q) 5 North America 2 South America 8 South America ~ p.. 4 South America 6 South America t 8 54 21 4o I>, r-i 0 p.. 3 Africa 4 Africa 4 Africa 3 Africa 5 Circum-Mediterranean 10 Circum-Mediterranear 12 Circum-Mediterranean 4 Circum-Mediterranean 5 East Eurasia 26 East Eurasia 12 East Eurasia 18 East Eurasia ~ 3 Insular Pacific 20 Insular Pacific fil 6 Insular Pacific 15 Insular Pacific ~ 9 North America 14 North America en 17 North America 6 North America ~ 8 South America 14 South America ~ 8 South America 12 South America 33 88 59 58 2 Q, =- -.319 p • -.153 X • 4.149 p<..05 C • .151 >.. • .013 C v-l 0 TABLE 8--Continued ( e) ( f) Authoritative Command Prestige for Warriors Present Absent High Moderate, Low ~ 7 Africa MO Africa 13 Africa 4 Africa 5 Circum-Mediterranean 4 Circum-Mediterranean 6 Circum-Mediterranean 4 Circum-Mediterranean 1 East Eurasia 1 East Eurasia 1 East Eurasia 1 East Eurasia 1 Insular Pacific 7 Insular Pacific 4 Insular Pacific 3 Insular Pacific MO North America 4 North America 6 North America 3 South America 7 South America 4 South America 4 South America 17 :/9 32 22 2 Africa 1 Africa 2 Africa 3 Africa 9 Circum-Mediterranean 4 Circum-Mediterranean 5 Circum-Mediterranean 11 Circum-Mediterranean 16 East Eurasia 8 East Eurasia 4 East Eurasia 22 East Eurasia 10 Insular Pacific 8 Insular Pacific ~ 7 Insular Pacific 13 Insular Pacific 4 North America ~5 North America Cl) rn 7 North America 16 North America 9 South America 7 South America ;! 8 South Am.erica 7 South America 50 43 33 ,_ 72 Q = • 521 /, • • 268 x2 ... 11 • 428 p ~ • 001 C • .222 X. • .104 C C • .259 X.c • .154 \J,J 0 N TABLE 8--Continued (g) (h) Elaborate Rewards for Warriors Violence Believed to Solve Problems Present Absent Yes No 13 Africa 2 Africa 9 Africa 6 Africa 5 Circum-Mediterranean 2 Circum-Mediterranean 8 Circum-Mediterranean 2 Circum-Mediterranean 4 Insular Pacific 1 East Eurasia 1 East Eurasia 2 East Eurasia 5 North .Alllerica 2 North America ~ 5 Insular Pacific 1 Insular Pacific (1) 10 South America 2 South America (I) 5 North America 2 North America (1) ~ 8 South America 1 South America ll. 33 9 t 36 14 » r-4 0 ll. 1 Africa 2 Africa 5 Africa 5 Circum-Mediterranean 6 Circum-Mediterranean 9 Circum-Mediterranean 5 Circum-Mediterranean 5 East Eurasia 8 East Eurasia 8 East Eurasia 14 East Eurasia 9 Insular Pacific 2 Insular Pacific 8 Insular Pacific 6 Insular Pacific 10 North America n1 North America 9 North America 10 North America 8 South America 4 South America 3 South America 8 South America 38 33 37 48 Q. = .522 /, • .251 x_2 ... 1 .091 p< .01 Q = .539 /, = .276 x2 a 10.275 p <..01 C • .243 C .... 266 A • .177 C \.JJ 0 \.JJ TABLE 8--Continued ( i) ( j) Felt Value of War Military Success Positive Negative Expanding Not Expanding 13 Africa 5 Africa 11 Africa 10 Africa 8 Circum-Mediterranean 1 Circum-Mediterranean 3 Circum-Mediterranean 6 Circum-Mediterranear .+) 2 East Eurasia 1 East Eurasia 2 East Eurasia i:: 3 Insular Pacific 1:: 3 Insular Pacific QI 1 Insular Pacific 7 Insular Pacific QI CT.I 0) 5 North America 4 North America QI QI 1 North America 8 North America S.. S.. p.. 6 South America 2 South America p.. 4 South America 6 South America 37 15 21 39 5 Africa 7 Africa MO Circum-Mediterranean 6 Circum-Mediterranean 6 Circum-Mediterranean 12 Circum-Mediter.raneat ., 7 East Eurasia t>o East Eurasia 4 East Eurasia 27 East Eurasia 10 Insular Pacific i:: 7 Insular Pacific 3 Insular Pacific 18 Insular Pacific QI 8 North America 114 North America 4 North America 0) 19 North America ~ 3 South America m2 South America 6 South America 15 South America 38 64 23 98 Q = .612 p = .321 x2 = 15.842 P < .001 2 Q. = .393 P - .176 x - 5.575 p< .02 C • • 305 Ac ,.. • 293 C • .173 TABLE 8--Continued (k) ( l) Military Expectation of Plunder Military Expectation of Trophies and Honors Present Absent Present Absent 18 Africa 2 Africa 9 Africa 11 Africa 10 Circum-Mediterranean 4 Circum-Mediterranean 6 Circum-Mediterranean 2 East Eurasia 1 East Eurasia 1 East Eurasia 2 East Eurasia 5 Insular Pacific 3 Insular Pacific ~ 3 Insular Pacific 5 Insular Pacific QI 7 North America 3 North America II) 7 North America 3 North ·America QI 8 South America 2 South America M 3 South America 7 South America A. t 50 11 t 27 34 >- >-M M 0 0 A. 3 Africa 3 Africa A. 3 Circum-Mediterranean 6 Africa 7 Circum-Mediterranean 10 Circum-Mediterranear 7 Insular Pacific 17 Circum-Mediterranean 10 East Eurasia 21 East Eurasia 10 North America 28 East Eurasia 12 Insular Pacific 9 Insular Pacific 5 South America 14 Insular Pacific 15 North America 4 North America 9 North America 12 South America 10 South America 17 South America 59 57 25 . 91 2 Q • .486 /, = .237 X • 9.938 p<. .01 C • .291 C • .231 Ar • .033 \..N 0 Vl 306 of the geographical representatives but not spe ci f ic i dent i - fi ca tions i n mos t cases . The regions are abbrevi ated as A (Africa), C (Ci rcum- Mediterranean), E Eastern Eurasia) , I (I nsular Pacific ), N (North America ), and S (South America) • . By inspection of the se numerous tables, one can see t hat societies from Africa are exerting consistently strong i nfluence upon the corre lations, followed by that fro m South Ame~cica a nd thirdly by some i nfluence fro m the Cir cum- Me di terranean region , specifically those societies on the African continent . Whatever the relationship of polygyny to the warfare variables, I cannot accept an explanat ion of its presence as t he bas i s f or Otterbein ' s fraterna l i nterest groups . It seems very tenuous to argue that the men are get- ting together because of their common re lationship to certain women . Patri l ocali ty as a s timulus for organj_zation · is not significant ; it does not correlate with patri- dominant labor . I ts s trongest assoc i ation i s with patrilineal descent , and each i s pr edictable from the other 84 per cent of the time . A large r category of "uniloca l " residence (patrilocal, viri- local , mat r ilocal , uxoriloca l , avunculocal) does have a nega- tive association with subjugation as a military expectat ion ( p < . 05) and a positive one with revenge ( p < . 05), a lthough a weak l ambda ( . 047) indicates tha t unilocal r esidence i s predictable from revenge but not the othe r way around . Patrilinea l descent , however , when i solated from all other forms of de s cent , does have a strong posit ive 307 relations hi p with the presence of i n t ernal war , s i gnificant at the . 01 level . Furthermore , patrilineali t y expl a i ns the presence of internal war 18 . 4 per cent of the t i me . Patri - lineality does not have a significant r elationship with external war- attacki ng . Appa re ntly it i s not res i dence a lone that can lead to i nternal war but the jural rule s of corpo r - ate descent groups t hat organi ze i ts members into a cooperating unit . A second worl d- wide confi guration revolves around the foo d- co l lecting communi ty . I f people depend upon food- colle c t i ng ( gathe ring , fishing , or hunt i ng ) for their primary means of livelihood , they will be living i n nomadic, semi- nomadic , or semi- sedentary se ttlements (predic t able 50 per cent of the time) averagi ng fewer than 100 pers ons although the popul ation can be as high as 400 ( both significant at hi gher than the . 001 l evel). Such a c ommunity may have up to 2 loca l jurisdictional l evels but no more than 1 jurisdic- t i ona l l evel beyond the loca l community and none at a ll 24 per cent of the time ( both s i gnificant at t he .. 00 1 l eve l) .. One may predict with accuracy of 18 . 8 per cent that those c ommunit i es with up to 400 residents do not have any super- ordinate political organi zation , and one 1 s accuracy of pre- dic ti on rises to 36 . 2 per cent i f the community has no more than 100 persons . While nomadic and semi- sedentary se t tle - ments may pos sess matrilineal ki n organization , the smaller communit i es are predict abl y bi l ateral, ambilateral , or 308 duolateral (p < . 05 ; A = . 128 ) . There i s no class strat i fi - cation i n such communit i es ( . 001) . As for assoc i ations with warf are , onl y communities of les s than 100 persons and food. - collectors generate s i gnifi cant negat i ve asso ciations with offens i ve external war ( < .02 and < . 05 res pe ctively) . Mobilization through age - sets , military soci et ie s , or standing armies is absent (.001) but the absence i s predictable only through knowledge of community size as 0- 400 persons (A= • 222 ). Anyone i n the communi ty can rr.ake the decis i on to go to war(< . 01 ) but it is not predict- a ble from knowle dge of settlement pattern , c ommunity size, or food-collecting . Low military expe ctations are consistently present although there is variation i n the signi f icance leve l s a nd predict ive s t rength between military expe ctations and each of the basic variables in the configurat i on . For ease i n presentation , I have arranged the correl at ions as follows . Military Expectations I (Low : defense , revenge , plunder ) with : nomadic , semi- sedentary settlements , <. 01 , pre- dict i ve value of . 085 ; community s i ze of 0-400 , < . 01 , predictive value O; communi ty s i ze of 0-1 00 , . 001 , pre- dictive capability of .1 62 ; food-collectors , < . 01, pre- dictive capability of . 155 . Military Expectat i ons II (Low : defense , revenge , plun- der , trophie s and honor) wit h : nomadic , semi- sedentary 309 settlements , < . 02, with a predictive value of . 086 ; community size of 0- 400 , < . 001 , predictive value . 210 ; coillIIluni ty size of 0-100 , < . 01 , predictive value . 136 ; food- collectors , < . 01, predictive val ue . 110 . There i s i nformal command i n war or no c ommand a t all ( p < . 001) , and it s s tronge s t pre dictor variable i s settle- ment pattern ( . 299 ) . The smaller communities of f ood- collectors have no elaborate rewards f or their warriors ( p < . 02-. 01), and subsistence i s the better predi ct or of such an absence ( . 11 9 ) . Communi t ies of 0- 400 do not expe c t vio l ence to solve the i r problems ( p < . 02 ) with a low pre- dictive value ( . 05 ) . Food-colle ctors consider violence and war a ne cessary evil or avo i d them (p < . 05), pre dictably ( . 107) . Nomadic and semi - sedentary communi t ies of 0-100 are not mi litary succe sses through expanding their terri tory ( p < . 05-. 02) . Subjugat ion as a mi litary expectation is sig- ni fi cant l y absent ( p < . 001 ) ; revenge i s present with the mode of sett l ement ( p < . 01) and with food- collectors ( . 001) . The att ribute food-collectors i s also positively associated with plunder at the . 05 l eve l. Communit ies of 0- 100 persons do not go to war for l and ( p < . 05 ) . None of the specific military expectations generate predictive direction . Ot;_t of all this , I have sele c ted a few tables so that the reader may i dentify t he member s ociet ie s in thi s config- uration , as in the first , and their geographical di stri but ion . 310 The di stri but ion i n the f i rst table (Tab l e 9) is : a b C d Africa 4 3 0 20 Circum- Mediterranean 0 7 0 21 Euras i a 7 4 0 23 I nsular Pacific 3 0 2 26 North America 16 6 5 6 South America 10 2 2 18 Sixty-five per cent of the mobile food-collectors are i n the New World , and nearly two- thirds of those are i n North Ame rica . I n the second t able (Table 10) inte rcorre lating nomadic or semi-nomadic s ettlements and a community popu- lation range of 0-400 persons , the geographical represen- tat ion i n t he cells is : a b C d Africa. 6 1 1 1 7 Circu.m- Me dite rranean 4 2 7 14 Eurasia 10 1 1 5 8 I nsular Pa cific 3 0 21 6 North America 18 3 7 4 South America 1 2 0 14 6 Th i s time , 58 . 5 per cent of the mobile communities of . less t han 400 population are in the New Wo r l d and 60 per cent of tho se a re i n North Ame r ica . Those marked with an asterisk have population of 100 or fewer . Table 11 shows the results of the correlation between subs i stence and offensive external war . The results are a significant negat ive association between foo d- collectors a nd 311 TABLE 9. INrERCORRELATION OF SEI'TLEMENT PATrERN AND SUBSISTENCE Food-Collectors Food-Producers Ingalik Comanche Lozi Hidatsa ~ Cp. Eskimo Chiricahua Ila Pawnee ~ Montagnais Warrau Masai Omaha Q) 0) 'O +r>-: Micmac Siriano Fulani Huron Q) Cl) Q) Saulteaux Nambicuara Somali Havasupai ..!i ffi damanese Slave Botocudo Bisharin Papago Elr.-p1 Q) edda Kaska Shavante +> Teda Goajiro aCl-) ~ inu Twana Aweikoma Rwala Timbira "M Gilyak E. Pomo Lengua Lapps 11 ukaghir Yokuts Abipon Kurd ! Chukchee Paiute Tehuelche Basseri Badjau Klamath Yahgan Toda 0) Tiwi Kutenai Kazak +> i: Aranda Gros Ventre 40 Khalka 22 Q) ~ Manus Thonga Gheg Tobelor Mundurucu r-1 +> Mbau Fijians +> Mbundu Romans Orokaiva Cubeo Q) Aleut Kongo Basques Kimam Cayapa Cl) Eyak Nyakyusa Irish Kapauku Jivaro ~ Q) Haida Luguru Russians Kwoma Amahuaca 'a. Bellacoola Kikuyu Abkhaz N. Ireland Inca ~ Yurok Ganda Annenians Trobriand Aymara 0. .. Miskito Nkundo Basseri Siuai Trumai 0) Callinago Banen Gon.d Tikopia Tupinamba +> Q) Santal Pentecost Cayua r-1 ~ u. Pradesh Ajie Mapuche :x.:. Burusho Maori Alorese Lalo Marquesan 0) 11 Lepcha Samoans Q) Garo Gi lberts +> D'J Serna Naga Marshal ls Q) ~ zande Bunnese Trukese :x: taro Nuba Palaung Yapese hilluk Vietnamese Palauans O'l +r>:: Rhade Ifugao Cl.I Khmer Itayal ~ Siamese Creek r-1 +> Semai Natchez +> Q) Nicobarese Zuni C/l Tanala Papago +> i: N. Sembilan Aztec Q) ~ Chinese Popoluca e ara Manchu Y. Maya Q) p. ubians Koreans Bribri El iffians Japanese Cuna H gyptians Javanese Haitians +> C, ebrews Balinese Yanomamo p«I. abylonia Iban Carib 8 9 ks Toradja Saramacca 114 c.., Q. = .917 Ii = = 69.265 p <.. .001 C = .522 "' C = .3. 67 "' r 312 TABLE 10 . INTER CORRELA TI ON OF SETI'LEME:t--J""T PATTERN AND MEAN COMMUNITY SIZE Mean Size of 0- 400 Persons Mean Size of 400 Persons or More , Nama Yukaghir* Pawnee Lozi Kung* Chukchee* Comanche Somali s:: Q) Ila Badjau Chiricahua* Kurd e Q) Hadza* Tiwi Havasupai Kazak rl Mbuti* Aranda* Papago Hidatsa ~ Cl) Masai Ingalik* Goajiro* Omaha Cf.) Fulani Cp. Eskimo* Warrau* Huron t- Teda* Montagnais Siriono* a:! Rwala* Micmac* Nambicuara* ~ Lapps* Slave* Timbira -Q) 'd Cl) Yurak Saulteaux Botocudo* Cf.) I Basseri Kaska* Shavante •rt Toda* Twana* Aweikoma* ffi Khalka* E. Pomo Lengua ~ Andaman* Paiute Abipon ·rt "d Vedda* Klamath* Tehuelche a:! 8 Ainu* Kutenai Yahgan* z Gilyak* G. Ventre 53 7 Thonga* Serna Naga Trukese Tiv Khmer Mbundu Palaung* Yapese Ibo Siamese Nyakyusa Semai* Palauans* Fon Negri Sembilan Luguru Nicobar Atayal Bambara Chinese Kikuyu Tanala Aleut Tallensi Javanese Ganda Manchu Eyak* Otoro Nuba Balinese Nkundo Koreans Haida Shilluk Kimam Banen Japanese Bellacoola* Songhai Tikopia Ashanti Toradja Yurok* Hausa Mbau Fijians Mende Iban* Creek Kanuri Ifugao Azande Alorese Huichol Konso Natchez ~lolof Orokaiva* Miskito Riffians Zuni Fur Kapauku Bribri* Egyptians Aztec Kafa* Kwoma Callinago Hebrews Popoluca Amhara Manus Yanomamo Babylonians Yucatec Maya Nubians* Lesu Carib* Turks Cuna Gheg Trobriand Saramacca Romans Haitians Abkhaz Siuai* Mundurucu Basques Cayapa Punjabi Pentecost* Cubeo Irish Aymara Gond* Ajie* Jivaro* Russians Tupinamba Santal Maori Amahuaca* Annenians Burusho Marque sans Inca u. Pradesh Lalo Samoans Trumai* Burmese [,epcha Gilberts Cayua* Vietnamese Garo Marshal ls Mapuche* Rhade 75 45 Q = .639 /J = .269 X- ~ = 12.994 C = .259 p< .001 * indicates societies whose average size is 100 or fewer persons. 313 TABLE 11. INTERCORRELATION OF SUBSI STENCE AND EXTERNAL WAR-A'Pl'ACKING - Continual/Frequent External War Infrequent Internal War Nama Miskito Kung Slave Yurak Samoyed Callinago Mbuti Kaska Ainu Nambicuara Andamanese Bella cool a O') J..t Yukaghir Botocudo Vedda Twana 0 -+' Chukchee Shavante Gilyak Yurek 0 (1) Aleut Aweikoma Badjau E. Pomo r-1 r-1 Micmac Lengua Tiwi Yokuts 0 0 Eyak Abipon Aranda Paiute ~ Haida Tehuelche Manus Kutenai 0 t£ Klamath Ingalik Warrau Gros Ventre Cp. Eskimo Siriano Comanche Montagnais Yahgan Chiricahua 22 Saulteaux 25 Thonga Romans Aztec Kongo Omaha Lozi Irish Yucatec Nyakyusa Havasupai Mbundu Russians Maya Luguru Huichol Ila Abkhaz Br ibr i Tallensi Cuna Kikuyu Kurd Yanomamo Songhai Goajiro Ganda Basseri Saramacca Nubians Haitians Banen Punjabi Mundurucu Turks Carib Tiv u. Pradesh Jivaro Gheg Cubeo Fon Burusho Inca Basques Cayapa Ashanti Kazak Aymara Lapps Trumai Mende Khalka Timbira Gond Cayua Bambara Lolo Tupinamba Toda Azande Garo Mapuche Santal ' Otero Nuba Serna Naga Lepcha O') J..t Shilluk Vietnamese Burmese (1) 0 Ingassana Khlner Palaung .g e Masai Siamese Rhade Wolof Japanese Semai P. Fulani ~ Javanese Nicobarese 0 Hausa Than Tanala 0 Ix. Kanuri Orokaiva Negri Sembilan Fur Pentecost Chinese Somali Ajie Koreans Amhara Marque sans Alorese Bisharin Trukese Kapauku / Teda ijidatsa Kwoma Tuareg Pawnee Trobrianders Riffians Huron Tikopia Egyptians Creek .Maori Hebrews Natchez Marshallese Babylonians Zuni Yapese Rwala Papago 76 Palauans 43 Q = -.335 /, = -.156 x2 = 4.053 c = .154 P = .05 .o41~ 314 continual or frequent offensive externa l war and , converse ly, a positiv e association between food- producers and continual or frequent offensive externa l war . Prediction is very weak , however ; knowl edge of subsistence correct i ng predict s exter- nal war only 4 . 4 per cent of the time . The geographical distribution of the soci eties by cells is : a b C d ·Africa 1 2 1 7 4 Ci rcurn- Mediterranean 0 0 20 6 East Euras ia 4 3 1 3 1 3 I nsular Pa cific 0 4 7 9 North America 8 1 3 8 3 South Ameri ca 9 3 1 1 8 Cell frequencies 22 25 76 43 166 Note that once aga i n 70 per cent of the 47 food-coll ecting societies are in the New World and two-thirds of those are in Horth Ameri ca . North America has consistent ly exerted the strongest i nf luence i n the food-collecting community configu- r ation . I shall have occasion to discuss the relat ionships of food- producers below . In the finest subsistence scal ing for food-collectors , signifi cant intercorrelations turned up that did not with the broad di s tinction between collectors a nd producers . Gather- ers are negat i ve ly associ ated with defensive external war (exte r nal war-being attacked ) (< . 01; A = . 094) ; they do ~o t go to war for plunder ( < . 05; A = .059) ; they do not go t o war for the lumped expectations of land and plunder 31 5 (< . 10 > . 05) . Fishers do not engage in offensive external war(< . 05 ; X = .074) ; nor do t hey go to war to subjugate people ( . 02) . They do go to war for plunder ( < . 05) but not predictably ; they do not go t o Kar f or the l umped expecta- t i ons of subjugation and tribute (< . 01 ) but not predi ctably . Hunters do go to war for plunder (< . 01) but n ot pre di ctably . Ten of t he 13 s ocie ties with t hi s associat ion a re i n t he New World (Montagnais , Mi cmac, Sl ave , Gros Ventre , Comanche , Si riano , Aweikoma , Lengua , Abi pon , and Tehuelche). Recall that the sub~istence desi gnat i on for all the socie t i es i s according to the manner i n which the peopl e deri ve most of their f ood ; there i s only 1 subs i stence i dent i f ication f or each society . Hunters do go to war for revenge ( < . 01) but not predictabl y . The thi rd configuration that I have i s ola t ed i n the 1-rnrld sample has t he l argest number of vari ables and is the most e l aborate i n i ts re l at i onships . Al t hough t here may be some awkwardness of fit , I call this the state confi gurat ion i ncluding petty paramount chiefdoms as well as small and l arge "states ." 0tterbein ' s variable of "centralized politi- cal organizat ion ," a s contras te d to "uncentrali zed ," woul d do a l so. Be caus e the combi nations of t he i nt ercorrel at i ons a re so num erous , I have arranged all those significant at the . 05 , .01, and . 00 1 l eve l s i n Table 12 , l eavi ng non- signifi cant i ntercorre l at i ons bl ank . Then I have i ncluded fourfold tables of several of t he more i nt erest i ng TABLE 12. SIGNIFICANCE LEVELS OF ASSOCIATIONS IN THE STATE CONFIGURATION Variable 8 9 12 13 14 15 19 22 23 24 25 30 31 32 36 37 38 33 35 45 8 9 .01 12 .05 13 . 001 14 .001 15 . 001 .001 .001 19 .001 .001 .001 .001 .001 22 .05 .01 .001 .05 .01 23 .05 . 001 24 .001 .001 .001 .001 .001 .05 25 .001 005 . 01 .01 .01 .001 .01 .01 30 .001 .001 . 001 .001 .01 . 001 .01 .001 .01 31 .01 . 01 .05 .05 .02 .05 .02 32 .001 .01 .001 .001 .001 .001 .05 .001 .001 ~001 .05 36 .05 .02 .01 .02 .01 .001 .·02 .02 .001 37 .05 . 01 .001 .05 .001 . 001 .02 .001 .001 .001 38 . 01 .001 .05 .01 .001 .02 .01 .001 .001 .01 .02 33 .01 .001 .01 35 .01 .01 .001 .001 .001 45 .001 .001 .001 .001 .01 .001 .01 .05 .01 .001 \.-"1 O"\ 317 correlations , to illustrate their geographical representa- tion . I have atteri,pted to bring out the most powerful rela- tionships that have predictive direction . The attributes present in this configuration are : compact permanent and complex settlements ; mean size of the typical community of 400 and l arger ; 3- 4 levels i n l ocal polit ical organization ; 1-4 levels of jurisdictional hierarchy beyond the l oca l c om- munity ; 2- 4 levels beyond the local community ; f ood- product ion ; social stratif ication ; continual or frequent external war- attack i ng and be ing attacked ; military organi- za~ion i n the form of age - sets military so cieties , and stand- i ng armi es ; de cisions to go to war made by an officia l or a council ; high military expe ctations of sub j ugation , tribute , land , and sometimes trophies and honors ; high casualties; authoritative command in batt le; violence seen as a soluti on to problems ; war given high value ; military suc cess define d as territorial expansion ; the abs ence of revenge as a mili- tary expectation ; prestige for warriors; and rewards for warriors . Table 13 shows the re sults of intercorrelating 2- 4 jurisdictional levels beyond the local community with the i nc i dence of offensive external war . The following i s a breakdown of the societies into geographica l region by ce l ls . 318 - TABLE 13. INTERCORRELATION OF LEVELS OF JURISDICTIONAL HIERARCHY BEYOND THE LOCAL COMMUNITY AND EXTERNAL WAR- ATTACKING _ Continual/Frequent External War Infrequent External War Thonga Amhara Serna Naga Nyakyusa Lozi Tuareg Vietnamese Songhai Mbundu Egyptians Khmer Turks Ganda Hebrews Khmer Gheg Albanians Ill Fon Babylonians Siamese Burmese r-1 Q) Ashanti Irish Japanese Negri Sembilan t Azande Russians Javanese Chinese ..:I Wolof Kurd Creek Koreans ~ I Hausa Punjabi Aztec Alorese C\J Kanuri U. Pradesh Saramacca Palauans Fur Kazak Inca Somali Khalka 34 10 Nama Micmac Kung Copper Eskimo Ila Eyak Kongo Montagnais Kikuyu Haida Luguru Saulteaux Banen Klamath Mbuti Slave Tiv Gros Ventre Tallensi Kaska Mende Hidatsa Nubians Bellacoola Bambara Pawnee Lapps Twana Otoro Nuba Huron Gond Yurek Shilluk Natchez Toda E. Pomo Ingassana Comanche Santal Yokuts Masai Chiricahua Lepcha Paiute Fulani Zuni Palaung Kutenai Bisharin Papago Rhade Omaha Teda Yucatec Maya Semai Havasupai r-1 Q) Riff'ians Miskito Nicobarese Huichol t Rwala Bribri Andamanese Cuna ..:I Yurak Samoyed Vedda Goajiro .... I Basseri Callinago Tanala Haitians 0 Burusho Yanomamo Gilyak Warrau Lolo Mundurucu Badjau Carib Garo Jivaro Tiwi Cubeo Ainu Nambicuara Aranda Cayapa Yukaghir Timbira Kapauku Siriano Chukchee Tupinamba Kwoma Trumai Iban Botocudo Manus Cayua Orokaiva Shavante Trobrianders Pentecost Aweikome. Tikopia Yahgan Ajie Lengua Maori Marque sans Abipon Marshallese Trukese Mapuche Yapese Aleut Tehuelche 61 Ingalik 57 Q. = 2 .521 /, = .231 X = 8.645 p <..01 C = .225 319 a b C d Africa 7 1 1 1 5 Circum-Me diterranean 1 3 3 5 2 East Eurasia 9 3 8 1 2 I nsular Pacific 1 2 6 1 1 North America 2 0 14 1 6 South Ameri ca 2 0 31 1 1 Cell frequencies 34 10 61 57 162 Note that the members of cell ~--their level codings trans- l atable into petty paramount chiefdoms , small states , and l arge states , and experiencing cont i nua l or frequent offen- sive external war--are in large majority from Africa, the Ci r cum- Medi t erranean , and East Eurasia . Thirty-eight per ce nt of t he societies are from the Circum- Mediterranean regi on and 85 per cent are from t he . Circum-Mediterranean , Africa , .and East Eurasia . African ki ngd.oIIis and modern states from Eurasia accentuate the influence of ancient and modern state s from the Circum-Medit erranean . While the intercorre- l ation i s s i gnificant at the . 01 level, it generates no pre- dic t ive direction . Nor do the se cent ralize d polities have a significant assoc i ation with external war- being attacked ( Tabl e 14) . Using a no ther scaling from the Ethnogra phic Atlas, howeve r, s tre r.gthens coefficients and produces predictability but I a m not certai n how to interpre t t he cost . I n this i ntercorrela tion , the politica l attribute is 1 to 4 l evels of jurj_sdictional hierarchy beyond the loca l community . It admits su ch African societie s as the Nama , the Ila , t he Tiv , 320 TABLE 14. INTERCORRELATION OF LEVELS OF JURISDICTIONAL HIERARCHY BEYOND THE LOCAL COMMUNITY AND EX'rERNAL WAR-BEING A'M'ACKED Continual/Frequent External War Infrequent External War Nyakyusa Kurd Thonga Balinese Banda U. Pradesh Lozi Alorese Nkundo Mongo Khalka Ashanti Palauans Fon Burmese Hausa U) Azande Vietnamese Kanuri ,-i Q) Wolof Khmer Amhara t Songhai Chinese Hebrews H Fur Koreans Turks .::r I Somali Japanese Irish - · (\J Tuareg Javanese Russians Egyptians Creek Punjabi Babylonians Aztec Serna Naga Gheg Albanians Saramacca Siamese Armenians Inca 28 Negri Sembilan 17 Nama Ajie Tupinamba Kung Slave Ila Marques ans Botocudo Kongo Kaska Hadza Marshallese Mbuti Twana Luguru Trukese Aweikoma Tallensi Yurok Kikuyu Ifugao Cayua Nubians E. Pomo Banen Aleut Abipon Lapps Yokuts Tiv Montagnais Tehuelche Gond Havasupai Ibo Micmac Mapuche Toda Huichol Bambara Eyak Santal Bribri Otoro Nuba Bellacoola Burusho Goajiro Shilluk Klamath Lepcha Haitians Masai Kutenai Palaung Warrau ,-i Fulani Hidatsa Nicobar Yanomamo Q) t Konso Pawnee Andaman Carib H Bisharin Huron Vedda Mundurucu .,... Teda Natchez Tanala Cayapa I Riffians Comanche Gilyak -0 Shavante Rwala Chiricahua Chukchee Yahgan Basseri Zuni Badjau Lolo Papago Tiwi Garo Yucatec Maya Aranda Rhade Miskito Kapauku Semai Callinago Kwoma Ainu Cubeo Tikopia Yukaghir Jivaro Maori Iban Amahuaca Yapese Orokaiva Siriano Ingalik Trobriand Trumai Cp. Eskimo Siuai Timbira 65 Saulteaux 47 Q = .087 i = .039 2 X = .233 n.s. C = .038 321 and the Masai , and Circum- Mediterranean ones such as the Riffians and the Somali (Table 15) . The geographical di stri- but ion of this intercorrelati on with the pre sence or absence of frequent exterEal war- attaciing by cell i s : a b C d Africa 1 5 2 3 4 Ci rcum- Mediterranean 18 3 0 2 East Eurasia 1 3 9 4 7 I nsular Pacific 3 8 4 5 North America 8 2 8 14 South America 5 1 14 1 0 Cell frequencies 62 25 33 42 162 While the association is now significant at the .001 level, the relative effe ct of these 3 Old World regions is lessened . Africa accounts for 29 per cent of the soci eties in cell~ ' ar it , together with the Circum- Mediterranean and East Eurasia , accounts for 74 per cent . However, the correlation possesses predictive direction , although odd i n one way : knowledge of the presence of 1- 4 jurisdictional levels allows prediction of the presence of external war-at tacking 13 . 4 per cent of the time . . However , knowledge of external war allows predict ion of these jurisdictional leve l s 22 . 7 per cent of the time. Probably the latter l ambda is reflecting age - sets and other supra-community organizations that are direct ly i nvolved in war . Moreover , this political attribute has a posit i ve association with external war - be i ng attacked , one of t he few such in the entire study (Table 16 ) . The probability leve l 322 '!'ABLE 15. INTERCORRELATION OF LEVELS OF JURISDICTIONAL HIERARCHY BEYOND THE LOCAL COMMUNITY AND EXTERNAL WAR-A'Pl'ACKING ~ontinual/Prequent External War Infrequent External War Nama Amhara Khmer Kongo Palauans Thonga Bisharin Siamese Nyakyusa Kutenai Lozi Teda Japanese Songhai Omaha Mbundu Tuareg Ainu Turks Cuna Ila Riffians Javanese Gheg Ganda Egyptians Ajie Maria Gond Tiv Hebrews Marques ans Toda Fon Babylonians Micmac Santal / Ashanti Rwala Eyak Lepcha . Ill r-1 Mende Irish Gros Ventre Burmese Q) t Bambara Russians Pawnee Tanala H Azande Kurd Huron Negri Sembilan .::t Otoro Nuba Basseri Creek Chinese ...'.. Shilluk Punjabi Natchez Koreans Masai u. Pradesh Aztec Alorese Wolof Burusho Yuca. Maya Kapauku Fulani Kazak Miskito Trobrianders Hausa Khalka Saramacca Tikopia Kanuri Garo Inca Maori Fur Serna Naga Tupinamba Marshallese Somali Vietnamese 62 Yapese 25 Kikuyu Muncurucu Kung Slave Banen Jivaro Luguru Kaska Ingassana Narnbicuara Mbuti Bellacoola Yurak Samoyed Timbira Tallensi Twana Lolo Botocudo Nubians Yurek Yukaghir Shavante Lapps E. Pomo Chukchee Aweikoma Palaung Yokuts Iban Lengua Rhade Paiute Orokaiva Abipon Semai Havasupai Pentecoot Mapuche Nicobarese Huichol Trukese Tehuelche Andamanese Goajiro Ill r-1 Aleut Vedda Haitians ~ Haida Gilyak Warrau H Klamath Badjau Carib 0 Hidatsa Tiwi Cubeo Comanche Aranda Cayapa Chiricahua Apache Kworna Siriano Zuni Manus Trumai Papago Ingalik Cayua Bribri Copper Eskimo Yahgan Callinago Montagnais Yanomarno 33 Saulteaux 42 2 Q = .519 p = .276 X C = .266 \ = .134 AC r = .227 323 TABLE 16. INTERCORRELATION OF LEVELS OP JURISDICTIONAL HIERARCHY BEYOND THE LOCAL COMMUNITY AND EXTERNAL WAR-BEING ATTACKED Continual/Frequent External War Infrequent External War Nama Teda Javanese Thonga Negri Sernbilan Ila Tuareg Trobriand Lozi Balinese Nyakyusa Riffians Ajie Kongo Alorese Ganda Egyptians Marques ans Ashanti Kapauku Nkundo Babylonians Marshallese Hausa Tikopia Tiv Rwala Micmac Kanuri Maori Ibo Gheg Eya.'l{ Amhara Yapese Fon Annenians Kutenai Hebrews Palauans 11} Bambara Kurd Pawnee Turks rl Q) Azande Basseri Huron Irish t Otoro Nuba U. Pradesh Creek Russians ~ Shilluk Khalka Natchez Punjabi .:::t I Masai Garo Aztec Gond T'" IWolof Burmese Yuca. Maya Toda - Songhai Vietnamese Miskito Santal Fulani Khmer Saramacca Burusho Fur Chinese Inca Lepcha Konso Koreans Ti.lpinamba Serna Naga Somali Japanese Siamese Bisharin Ainu 58 Tanala 28 Hadza Jivaro Mbuti Huichol Luguru Amhuaca Tallensi Bribri Kikuyu Siriono Nubians Goajiro Banen Trurnai Lapps Haitians Lolo Timbira Palaung Warrau Rhade Botocudo Nicobarese Yanomamo Semai Aweikoma Andamanese Carib Yukaghir Cayua Vedda Mundurucu Iban Abipon Gilyak Cayapa Orokaiva Mapuche Chukchee Shavante Siuai Tehuelche Badjau Yahgan Cl) Trukese Tiwi rl Q) Ifugao Aranda ~ Aleut Kwoma ~ Montagnais Ingalik 0 Bellacoola Copper Eskimo Klamath Saul tea.we Hidatsa Slave Comanche Kaska Chiricahua Apache Twana Zuni Yurok Papago E. Pomo Callinago Yokuts Cubeo . 35 Havasupai 36 i :c: .184 x2 = 5.303 p<.05 C = .181 Ae = .016 Ar = .113 324 i s on.ly . 05 , but the l ambda i Ldicate , albeit weakly , that the poli t ica l attribute is predictable from the presence of defensive war 11 . 3 per c~nt of the t i ffie . There i s something else i nteresting happening here . Compar ing the s i gni f i cance level s obtained by 1-4 jurisdic- t i onal l evels to those by 2-4 jurisdictional levels , one f i nds the l evels reached by the lat t er either non- s i gnif i cant or slightly hi gher . I thi nk that the androcentri c configura- tion of t ribal soci eties , part i cularl y African ones , i s over- l appi ng the state configuration domi nat ed by the Circurn- Me diterranean region . This explains the assoc i ations achieved wi th the prest i ge and reward vari ables . Certai nly it is a thoroughly documented and c ommonly experi enced char- acteristic of t he state in Weste rn history that warfare is highly valued , particul a rly as the means t o so l ve poli t ica l a nd interna tional problems . Furthermore , withi n the state i ndi vidual fighters do not re ceive the publi c prest i ge and a ccolades tha t they do in militarist ic tri bal s ocieties . A military i nst i tution might, such as the Pentagon or the German General Staff . As I have d iscussed before , the stat e does not need i ncentives when it has legal obligation and force . But it is not the presence of s tatehood alone that elirni nates individual prestige and. rewards . Recall t hat i n the Eurasian region , which includes tribal societies as well as modern and ancient states , prestige for warriors , ~ewards for warriors , and war as a positively val ued phenomenon are 325 a sent . I contend that this is :further indication that it is the nature of the Circum- Meaiterranean utate t hat is domi - nating the world . I n Figure 1 , I have attempted to summarize diagram- matically the highest values of lambda and their direction among the variabl es in the state c onfiguration that generate any predictability a t all. The var j_ables are ident ified by their numbers ; refe r to Appendi x B for clarification , i f ne cessary . Many more correlations are stat i stically signi f i- can only, e .g ., the one between food- producers and external war in Table 8 ind others following i n Tabl e 17 . The l ambda val ues i n Fi gure generally are in the 20- 40 percent ile , which is only i n the moderate range , yet these are the strongest that appear i n the world sample . The patterning in the configurat ion I i nterpret as follows . There are 3 foci of 11 expl anation"--attri butes upon which most of the other vari- ables are dependent for any predi ct ive direction . They are the state l 13 ), external war- attacking (22) , and military success defined as territorial expansion of the c ul tural unit (38) . There are a lso 3 points that seem to be the most dependent upoL other variable s . They are c ompact permanent or complex sett l ements (8) , the belief in violence as a so l u- tion to problems (36) , and war against noL- members of the group as highly valued (37) . There is no predictive dire c- t ion between the state and external war , although the rela- ionship is significant at the . 01 level . The attributes 326 KEY Q Variables Most Often Independent fl'::\ Variables Most Fig. V::::::::J Often Dependent 1 The World: direction and strength of lambdas, focused on 2-4 levels of polity (13). 0 Other Variables Direction of Dependence and Strength of Predicta bility 327 TABLE 17 . SIGNIFICANT CORRELATIONS WI TH O LAMBDAS Pairs of Variables Phi p Compact , compl ex sett l ements (8 ) wi th : 3- 4 leve l s of local jurisdiction ( 12 ) . 1 50 <. 0 5 2-4 leve l s beyond the l ocal communi ty (1 3 ) wi th : Fo od- producers (1 5) , 354 . 001 Exte r nal war- a t tacki ng ( 22 ) . 231 <. 01 Of fici a l de cis ion to go t o war ( 25 ) . 232 <.01 Peace ceremony absent ( 28 ) . 273 <. 02 Military expe ctations of subjugat i on , tribut e, l and , t rophies and honors ( 29 ) . 258 . 001 Food- producers ( 15 ) wi th : Class s trat i fi ca t ion (1 9 ) . 320 . 001 Mobilization by age- sets ... standi ng armi es (24 ) . 298 . 001 Officia l de cis i on t o go to war ( 25 ) . 26 0 <.01 Class s t rat i fi cat i on (1 9 ) wi th : External war- attacking ( 22 ) . 228 <. 01 Mi l i tary success ( terri tori a l expansion)( 38 ) . 205 <. 01 Frequent external war-attacking ( 22 ) wi th : Mobilizat i on by age - sets ... standi ng armi es (24 ) . 177 <,05 Frequent exter nal war- be i ng attacke d (23 ) wi th : Absent mili tary success (expans ion , r eplaci ng, unchangi ng) ( 39 ) . 24 6 <. 01 Mobilization by age - sets . .. standing armies ( 24) with : Mi l itary expectations of subjugat i on , tri bute , l and , t rophies and honors (29 ) . 240 <. 01 Officia l decis i on to go to war ( 25 ) with : Milj_tary expectations of subjugat i on , t ribute , l and , trophies and honors (29 ) . 201 <. 02 Military success ( territorial expansi on )( 38 ) . 245 <. 01 328 TABLE 17 (Continued ) Pa irs of Variable s Phi p High pre s tige for warriors (33) wi t h : Milit a ry success (expansion, replacing , unchanging) ( 39) • 1 96 < . 02 Rewar ds for wa rriors (35) with : Trophi es and honors (44 ) . 341 . 00 1 generat i ng the strongest predict ions a re cont i nua~ or fre - quent external war - attacking (22 ) , followed in strength by the s tate (13) . Military success ( 38 ) is only a weak pre- dictor . In looking a t Figure 1 as a whole , several a ttributes of the androcentric conf i gurat io ri that I commented upon earli er as i mpingi ng upon the state configuration are clus- tered i n the lower half of the diagram , while the state and i ts depe ndent attri butes are clus tered in the upper half . Midway are the attributes of belief i n violence as a solut i on to problems (36) and war as highly valued (37) . They may be a point of transition between t he 2 configurations , most likely to be predicted through knowledge of offensive exter- nal war but a l s o linked to the state . On the world s cene the se a ttitude variables are de pendent , not i ndependent as I have generally hypothes~zed. Looking at the bottom of Figure 1 , notice t he re lative ly strong predict i ve 329 capabili ty generated by the attributes of high prestige for warriors (33) and elaborate rewarQs for warriors (35) , and the nearly balance d influence of external war-attacking (22) and external war- being attacked (23) upon each other . Read- ing the diagram as a map , I trace the following connections . I f political communities of a culture experi ence attack from outs i ders , they may then wage offens ive war . Note that attacking offensively i s not the primary condition . If warriors are bestowed with elaborate rewards , t hen being a warrior i s predi ctably a position of high prestige in the community . The existence of such prestige gives war itself an activity of high value and makes very likely the be lief that war solves problems . That in turn provi des a moderate prediction that high casualt ies will be sustained . Military success is only a weak predictor of mi litarist values and attitudes . Once there exist 2- 4 levels of poli tical organi- zation i n a community, however, there is the relatively strong possi bility that there will be class stratification i n the commw1i ty , authoritative military .command , standing armies , and military expectations of subjugation , t ribute , and land, and relatively moderate to weak predict i ons that attitudes toward violence and war are posit i ve . I would like to suggest that the patterning in the diagram refle cts , how- ever faintly , the geographica l distribut i on of configurations in the Circurn-Me diterranean region and in the sub- Saharan African region . 330 Figure 2 shows the relationships among the variables when the attribute of jurisdictional l eve l s i s expanded fr om 2- 4 levels to 1- 4 levels . Now there is a direct association between polity and offensive war , where the type of warfare predicts polity almost 23 per cent of the time . There is a l so a direct association between polity and high military expectations, and the predictive directions of the relation- ships between jurisdictional levels and military organization, and jurisdictional l evels and class stratification , are oppo- site to those in Figure 1. Internal war becomes part of this configuration through its relationship to unilineal des cent , which in turn is re l ated to polity . I nternal war becomes tie d into each configuration through a different at t ribut e : for t he world- wide intercorrelations i t is descent groups ; for Afri ca it i s the absence of authoritari an military l eaders ; for the Circum-Mediterranean region it i s personal profit and prestige ; for North Ameri ca it is compl ex settle- ments ; for South America i t is vio lence as an expe c ted solu- tion to problems and patri l oca lity . The only significant associations with int ernal war in the East Eurasian region are with endogamy and elaborate rewards for warriors . Neither of these attributes ties internal war into the larger regiona l co~figuration . The re are no significant associ- at ions with interna l war within the Insular Pacific region , although internal war is significant ly associated with this r egion . 331 ' Fig. 2 The World: direction and strength of lambdas, focused on 1 -4 levels of polity (14). 332 The 11 map " is mult i plied i n complexity , however , because several of the attributes are predictable i n 2 direc- tions . My i nclusion of only the strongest lambda values and their di rections obscures what I can only l abel as feed back relationships among the attributes , a l though one may have stronger i nfl uence than the other . Nevertheless , a feedback or circul ar relat i onship exists . For instance , one can pre- dict from knowle dge of high prestige for warriors t hat war i tself will be highly val ued 49 per cent of the time . But from knowledge that war i s a highl y ~al ued a ctivity, one can predict 39 per cent of -the time that warri ors will be a pres- tigious or mandatory role for every mal e . That while the presence of rewards explains prestige 50 per cent of the time , the reverse is the case 34 per cent of the time. Not all of the relat i onships are two- way , however . Many are predictable in only one direction . In Table 18 , I i nclude the val ues for both directions of l ambda . I n further support of the conf i gurational argument , I include more substantive data i n Tables 19- 21. illustrating the relationships between t he state and military organiza- tion , military expe ctat ions , and military success , breaking the cell frequencies down i nto their geographical constitu- ents . The effect of societies from the Circum- Mediterranean ~nd part of the Eurasian region is apparent in all of them . Note , however , that the corre l at ions i n Tables 20 and 21 have no predictive direction . 333 TABLE 18 . SIGNI FICANT CORRELATIONS WITH LAMBDAS I N TWO DIRECT IONS Pai rs of Variables cp p "C Row attri bute Complex Sett l ements (8) with : Class stratification (1 9) . 407 . 001 -338 . 12 1 Age- sets ... standing a rmie s (24) . 392 . 001 . 354 . 1 77 Military expecta~ions II (30) . 323 . 001 . 280 . 272 High casualties (31) -351 <. 01 . 341 . 250 Authoritative command (32) . 4 31 . 001 . 408 . 373 Expe ctations of violence (36) . ,17 3 <.05 . 067 . 097 High value of war ( 37) . 1 75 <.05 . 11 3 . 149 Row attribute 2-4 Political Levels (13) with : Class stratification (19) . 567 . 001 . 289 . 439 Age - sets ... s tandi ng armies (24) .467 . 001 . 188 -339 High casual t i es ( 31 ) . 246 <. 01 . 037 . 257 Authoritative command (32) . 450 . 001 . 267 . 468 Ro~ attri bute 1-4 Levels Beyond the Local Community (1 4) with : Complex settlements (8) . 424 . 001 . 354 . 346 Mean community s i ze 0-100 (10) . 419 . 001 . 362 . 073 Uni lineal descent (17) . 245 . 001 . 171 . 209 Class stratification (19) . 529 . 001 . 432 . 193 Frequent external war- attacking (2 2) . 276 . 001 . 227 .1 34 Military expectations I (29) . 436 . 001 . 365 .31 9 Military expectations II (30) . 407 . 001 . 297 . 342 Authoritative command (32) . 450 . 001 . 300 . 323 Pr e s tige for warriors (33) . 209 . 01 . 125 . 031 High value for violence/war (37) . 323 .001 . 271 . 301 Row attribute Class Stratification (1 9) with : Military expe c tations II (30 ) . 355 . 001 . 086 . 293 Authoritative command (32) . 502 . 001 . 370 . 44 3 334 TABLE 18 (Continued ) Pairs of Variable s cp p ).. )._ r C Row attribute Frequent External War- Attacking (22 ) with : External war-being attacked (23) . 500 . 001 . 422 . 413 Military expectations I (29) . 346 . 001 . 1 94 . 21 9 Prestige for warriors (33) . 306 . 001 . 085 . 129 Elaborate rewards for warrior ( 35) . 502 . 001 . 359 . 405 Expectations of violence ( 36 ) . 455 . 001 . 346 . 414 High value for violence/war (37) . 453 . 001 . 355 . 429 Row attribute Frequent External War- Being Attacked (23) with : Rewards for warriors (35) . 397 . 001 . 194 . 256 Expectations of violence (36) . 21 7 <. 02 . 04 1 . 1 30 Row attribute Age-sets ... Standing Armies (24) with : Military expe ctations II (30) . 292 . 001 . 065 . 275 Authoritative command (32 ) . 380 . 001 . 24 1 . 31 3 Row attri bute High Military Expectations II (subjugat ion, tribute, land) (30) wi h : High casualt ie s (31) . 269 <. 02 . 211 . 167 Authoritative command (3 2 ) . 275 . 001 . 260 . 194 Row attribute High Casualties (31) with : Authoritative command (32) . 258 <. 05 .152 . 152 Expectations of violence ( 36) .297 < . 02 . 226 . 040 Row attribute High Prestige for Warrior ( 33 ) with : Elaborate rewards for warrior ( 35 ) . 531 . 00 1 . 500 . 34 1 Expectations of violence ( 36 ) . 486 . 001 . 333 . 393 Highly valued war ( 37) . 503 .001 . 390 . 493 Row attribute Elaborate lie,ards for Warriors ( 35 ) with : High value for war ( 37) . 379 . 001 . 220 . 319 335 TABLE 19. INTF..RCORRELATION OF MILITARY ORGANIZATION AND POLITICAL ORGANIZA'J'ION Mobilization Age Sets ••• Armies Informal or Absent 7 Africa 2 Africa ·12 Circum-Mediterranean 4 Circum-Mediterranean 10 East Eurasia 3 East Eurasia 3 Insular Pacific 3 Insular Pacific 2 South America 1 North America 1 South America 14 s::::$ 0 ·r-t •r-t s:::: ,-; +' ;::s Ill CJ § C) •r-t I:: 'd 5 Africa 12 Africa •r-t 0 +' ~) 0 1 Circum-Mediterranean 6 Circum-Mediterranean ·r-t ·r-l ,-; $.., ,-; :::1 2 East Eurasia 14 East Eurasia 0 Ill p.. I--, C) 2 Insular Pacific 20 Insular Pacific 0 t,..i,..:i 9 North America 22 North America 0 (j) ,-; ..G 6 South America 16 South America (j) +> > Q) 'd ,..:i S:::: 0 .- l>, 25 91 I (l) 0~ 164 Q. = .797 /, = .467 2 X = 35. 799 p <...001 C = .423 1c = 0 339 A. C r = .188 336 TABLE 20 . INTERCORRELATION OF MILITARY EXPECTATIONS AND POLITICAL ORGANIZA'I'ION Military Expectations Subjugation, Tribute, Land Other rl [/} +:> rl 6 Africa 3 Africa Q) § sro +r>o 13 Circum-Mediterranean 4 Circum-Mediterranean 0 Cl.). +> 10 East Eurasia 3 East Eurasia § CfJ 4 Insular Pacific O'l 2 Insular Pacific ~ QJ ell E bO 1 North America 1 North Ameri ca ~ 0 ~ i:: -o ro 0 +>>, Ct-i ...:I 2 South America ·rl Q) +> +> ·rl 'O t\l ~QJ u..c: § N ·rl 36 13 ~ b.O ~ i:: .t'0 0 § ·rl r-l +> ell C) -~ ~ 6 Africa ·rl "O 0 11 Africa +> (/J u 6 Circum-Mediterranean ·rl 8 East Eurasia 1 Circum-Mediterranean ·rl rl 0 s~ 9 Insular Pacific 11 Eas t Eurasia ~ ', C) 0 r+-i....:I 9 North America 15 Insular Pacific 0 5 South America 18 North America Q) rl ..c: 23 South America Q) +> :> Q) "O ...;i i:: 0 43 79 ..- >, I QJ 01:Q 171 Q; = .570 ,5 = .258 X2 ::: 11 • 349 p < • 001 C = .21~9 337 TABLE 21 • INTF.RCORR.r.:L.ATION OF MILITA..11.Y SUCCESS AND POLITICAL ORGANIZATION Military Success Territorial Expansion Net Expansion Absent rl U) +>rl 7 Africa 2 Africa Q) s::: scd +> 8 Circum-Mediterranean 9 Circum-Mediterranean ~ 0 Ci.). + c>d 3 East Eurasia 10 East Eurasia @ Ci) H 1 Insular Pacific 5 Insular Pacific Ill Q) cd § bl) 1 North America 1 North America p., 'Cl ~ s::: 2 South America >, Q) ·rl +> •rl 'Cl +> Q) .t:1 s::: ro p., 0 ro ~ 22 27 ·rl C: (1j s:::$ bO 0 ·rl H ·rl C: 0 +> ~ rl ro ~i 3 Africa 15 Africa 0 1/lt.) 2 Circum-Mediterranean 8 Circum-Mediterranean •rl ·rl +> ~~ 1 East Eurasia 17 East Eurasia ·rl rl .. 0 3 Insular Pacific 20 Insular Pacific p0., 0 Q) 'Cl ,...:i C: 0 21 106 ,- >, f Q) o Ill 175 Q == .609 /, = .296 X2 = 15.406 p <.001 C = .284 >.. r = .020 338 Going back to Naroll' s and Otterbein ' s f i ndings , a l though they are much more i nt erested i n military st rat egy and tacti cs than I, there are a f ew replications of their intercorrelat ions i n my research . Naroll (1969) finds that states with greater poli t ica l centrali zat i on are more like ly than those with l ess centralizat i or. to go to war . My find- i ngs support an argument for an opposite trend (Table s 13 and 15), that i n t he absence of supra-communi ty but nonstate organi zations , offens i ve ext ernal war is le ss like l y , i .e., greater centralization , l ess offens i ve war . Specifically, I have argued that there is some pattern of i nterveni ng var i- ables betwee n the "state " and "offensi ve war ." I have not i nc l uded the mode of succession of state rul ers as a variabl e , as Naro l l ha s , but I do i nclude t he mode of successi on of r ul ers of the loca l community, dichoto- mi zed as hereditary and nonhe r editary . While Naroll (1 969 ) fi nds t hat hereditary rulers are more likely t o be invo lved i n war than nonhere dita ry ones , I f i nd nothi ng s i gni f icant with regard to t he frequency of war of any type and local rul ers . There are , however , significant relationshi ps between heredi tary loca l rul ers and communi ties of 100-400 persons , politica l organization of 1 t o 4 l evels beyond the local community , unilineality , class stratificat i on , and-- where war occurs--wi tn official de cis ion , commencement by announcement or agreement , authoritative command , ano. the specific mi litary expectation of pl under . These 339 relationships are present ed i n Table 22a- h , once again broken down by geographical region . These results are curious . I n all 8 i nte rcorrelations hereditary succession does not generate predictive capa- bility, except in it s relat ionship to unili neal desce n t (Table 22c) , where the 2 attri butes are moderate predictors e qua l ly of each other ' s presence . Otherwise, the mode of s u ccession i s moderately predictable only as a dependent variable, whether upon community s i ze , class stratificat ion, authoritative command , or p l under . Of a ll 8 , the strongest pre dictor is official decision-making in g oirig to war . I cannot discern any ccnsistent geographical patterning among these variables , except that North America seems to l oad most often i n the common absence cell , and the cultures of the I nsular Pacific are appearing in the common presence cell in l arger numbers than they have in most of the other correla- tions presented so far . The overall picture , as far as I can see , is that on a world- wide ba sis local hereditary rulers exist at the will of the comrnunity--the Insular Pacific or African or New World headman who claims the right to be head- man through lineage membership but who is Harris ' hard- working paragon leading by example and reward, including plunde r in wa r, and n ot by force . I do not have the data to go beyond this to find any transition to an autocratic ruler , unless one wishes to interpret the predictive dependence of l ocal succession upon t he prese n ce of authoritat ive command TABLE 22. INTERCORRELA'l1IONS OF EIGHT OTHER VARIABLES WITH MODE OF SUCCESSION OF LOCAL HEADM.AN ( a) (b) Succession of Local Headman Succession of Local Headman Hereditary Non-Hereditary Hereditary Non-Hereditary p •rl C/J ~ 9 Africa 4 Africa 11 Africa 8 Africa 0 U) 4 Circum-Medit erranean 2 Circum-Mediterranear I 10 Circum-Medi terranea1 8 Circum-Mediterranean ~ (I) 6 East Eurasia 5 East Eurasia u Cl) 9 East Eurasia 10 East Eurasia P-< 9 Insular Pacific 7 Insular Pacific r-i rl crj Q) 1 1 Insular Pacific 6 Insular Pacific 0 8 North America 6 North America (.) > 0 ,~ .µ I 5 South America 3 South America HH 2 South America 1 South America •rl 0 OJ.::t- ii::: 0... .. ...µc: ...I.. 0 41 27 49 37 0 C+-i 0 (!.) ti) N (l) •r-l N 5 Africa 7 Africa 5 Africa 3 Africa Cl) ·rl Cl) 6 Circum-Mediterranean 7 Circum-Mediterranear 1 Circum-Mediterranean i:: crj ~ 7 East Eurasia 11 East Eurasia 4 East Eurasia 5 East Eurasia (I) (l) ~ ...µc: 6 Insular Pacific 6 Insular Pacific 4 Insular Pacific 8 Insular Pacific 0 5 North America 13 North America 8 North America 15 North America rl 7 South America 10 South America 10 South America 12 South America ,-( ~ Cl} 36 54 r-i Q) 31 44 > 0) H Q, = .390 p = 2 2 .201 X = 6.385 p <.02 Q-- .305 ~=.156 X = 3.922 P= .05 C = .197 >-. = .182 Ar . = .074 C = .154 A = .150 >.. r = .093 C C TABLE 22--Continued ( c) ( d) Succession of Local Headman Succession of Local Headman Hereditarv. Non-Heredit ary Hereditary Non-Her editary .µ s:: 11 Africa Cl) 8 Africa 8 Africa 3 Africa CJ CJ) 8 Circum-Mediterranean 6 Circum-Mediterranean 7 Circum-Mediterranean 3 Circum- Mediterranean (l) 11 East Eurasia 7 East Eurasia 2 East Eurasia 6 East Eurasia A 8 Insular Pacific 6 Insular Pacific t 2 Insular Pacific r-1 9 Insular Pacific Cl) cj (1) 8 North America 3 North America s:: C/l 4 North America 2 North Ameri ca s:: (1) South America 0 r-. 1 •rl 5 3 South America South America •rl p.., r-i ·rl g ~ CJ ·rl 51 33 , 5 Africa 3 Africa 1 Africa 8 Africa 0 ~ •r-1 ro 5 Circum-Mediterranean 2 Circum-Mediterranear § t 2 •rl .. r = .100 \...N ~ 1...0 350 TABLE 25. INTERCORRELATION OF EXTERNAL WAR-ATI'ACKING M1D EXTF..fu'JJ\L WAR-BEING ATI'ACKED Continual/ Frequent War-Being Attacked Infrequent War-Being Attacked Nama Abkhaz Huron Thonga Ila Kurd Creek Lozi Kikuyu Basseri Natchez Ashanti Ganda u. Pradesh Comanche Hausa Banen Khalka Chiricahua Kanuri Tiv Lolo Zuni Amhara Fon Garo Papago Hebrews Barnbara Vietnamese Aztec Irish Azande Khmer Yuca. Maya Russians Otoro Nuba Japanese Miskito Punjabi Shilluk Ainu .Callinago Burusho Masai Yukaghir Saramacca Senta Naga Wolof Javanese Jivaro Siamese Pulani Iban Inca Chukchee Fur Orokaiva Aymara Bribri Somali Ajie Timbira Yanomamo Bisharin Marquesans Tupinamba Mundurucu Teda Trukese Botocudo Shavante Tuareg Aleut Aweikoma Riffians Micma c Abipon Egyptians Eyak. Mapuche Babylonia Klamath Tehuelche Rwala Hidatsa Romans Pawnee 70 18 Nyak.yusa Kung Tiwi Hai ti ans Luguru Kongo Aranda Warrau Songhai Mbuti Kapauku. Carib Gheg Albanians Tallensi Kwoma Cayapa Basques Nubians Tikopia Yahgan Burmese Turks Maori Rhade Lapps Yapese Semai Gond Palau.ans Chinese Toda Ingalik Koreans Santal Copper Eskimo Trobrianders Lepcha Saulteaux Marshallese Palaung Slave Montagnais Nicobarese Kaska Bellacoola Andamanese Twana Ku.tenai Vedda Yurok Cubeo Tanala E. Por.10 Siriano Negri Sem. Yokuts Trumai Gilyak Havasupai Cayua Badja.u Huichol 19 Alorese Goa jiro 45 Q = .804 p = 2 .500 pz.001 X = 37.953 C= . 4471c = .413 Ar = .422 C 35 1 appear in thi s cell even though one cannot really say they have had. the problems of , say , the Songhai or the Trumai, but throughout i ts history , China has been attacked by woul d- be conquerors , and i ndeed conquered , but because of i ts remark- abl e historical and cultural capacity to absorb outsiders , China obvi ously has survived . The geographica l breakdown by cell s for Tabl e 25 i s : a b C d Africa 1 2 3 2 4 Circum-Mediterranean 14 6 3 · 3 East Euras i a 10 5 5 11 I nsular Pacific 6 0 2 10 North Ameri ca 14 0 3 1 1 South America 14 4 4 6 Otterbe i n has a l s o intercorre l ated off ensi ve and defensive external war , fi ndi ng a s i gni fi cant pos itive rela- tionship , but one that is les s strong t han i n my sample ( Tabl e 25 ) . I have figure d the l ambda val ues for his table, and they are a l so l ower . Moreover , a ccordi ng to the direc- tion of his l ambdas, the frequency of being attacked is more like l y to be explai ned by the frequency .of atta cking t han vi ce versa , whi l e I found the 2 vari able s to be about equal i n t heir predictive dependency , with a s light edge to be ing attacked explai ning at t a cking . Frequency of Being At tacked Continual or Frequent I nfrequent Continual 18 9 27 Frequency of or Frequent Attacking I nfrequent 5 1 5 20 23 24 47 cp = . 41 x2 = 7 . 98 p < . 01 ).. r = . 30 ).. . 3913 C = (adapted from Otterbein 1968 : 286 ) From his coding i nformation, I have been abl e to i dent i fy t he cultura l units occupying each cell (Table 26) , and they break down geographi cally as f ollows . a b C d Africa 2 3 1 3 9 Circum- Mediterranean 1 1 2 0 4 East Eurasia 2 2 0 3 7 I nsular Pacific 1 1 0 6 8 North Ameri ca 6 1 1 2 10 South America 6 1 1 1 -2 18 9 5 1 5 47 Wit hin this comparison are also i nstances of disagreement i n codi ng between Otterbein and me , i . e ., the Tiv , the Orokaiva , and the Marshalle se . I n re l ating my results further to those of Ot terbein , recall my argument earlier i n this chapter t hat polygyny as Otterbein has i so l ated it may be a misleading variable . Cor- relation of who decides to go to war ( "an official " or "any- one " ) with the frequency of internal war is non- significant 353 TABLE 26. RECONSTRUCTION OF OTI'ERBEIN1S INTERCORRELATION OF EXTF.RNAL WAR-A'I'I'ACKING AND EXTERNAL WAR-BEING ATTACKED Continual/ Frequent War-Being Attacked Infrequent War-Being Attacked Ila Mossi Nandi Tiv Ancient Egyptians Ingassana Yukaghir Somali Thai ( 1 600 A. D. ) Tibetans Kurtachi Serna Naga Comox J avanese (1 300 A.D. ) Wishram Aztec Plains Cree Mundurucu Fox Cherokee Papago ~ (I) Saramacca g, Jivaro (I) H Aymara µ:., Tehuelche "rl :::l Abipon cj s:: Timbira ·rl +' s:: 0 u 18 9 Gisu . Dorobo Albanians ( Gheg) Ambo Mutair Amba Santa Ana J apan ( 1200 A. D. ) Trumai Toda Andamanese Tiwi Orokaiva Marshal lese Lau Tikopia Hawaiians Copper Eskimo Monachi Motilion 5 15 A = .3913 C 354 in the ODYSSEY , contrary to Ot terbein (1968) . Comparing the results of OtterbeiL 1 s " centralized political systems " wi th "initiating party " to my "2-4 jurisQicti or.al leve l s beyond the l ocal communi ty" with "decision to go to war ," the ODYSSEY produces something i nteresting . I have summarized the i nte r- correlations i n t abular form and computed l ambda f or Otter- be i n ' s data . I nitiating Party Anyone Offi cial Centralized polit ical systems 2 14 1 6 Uncentralize d political 16 10 26 systems 18 24 42 cp = . 48 x2 = 9 . 73 p < . 01 [ A = . 33 ; Ar . 22 ] C = (Ott erbe i n 1968 : 282 ) Decision by Off icia l De cis ion or Council by Anyone 2- 4 levels beyond the lo cal 42 6 48 community 0- 1 leve l beyond the local 60 31 91 cormnuni ty 102 37 139 cp = . 232 x2 = 7.482 p < . 01 A = 0 Notice that while the level of significance is the s ame i n both intercorrel.ations , the moderate l ambda va l ues that one can compute out of Otterbe i n 1 s table di sappear i n the l a r ger run . 355 Next , recall that Otterbein (196 8 : 285) attempts to test the amity- enmity complex : the more frequent ly poli t ica l corrLmunities fight those who are culturally dissimilar , the l ess likely they are to fight political communi ties that are culturally s i milar . The resul ts of his i ntercorrelat i ons a re not s i gni fi cant . MiLe are , a l though not in the direction that Otterbe i n would wi sh . The ODYSSEY produces a positive association between frequent i nternal war and frequent offen- sive external war, while t hat between i nt ernal war and exter- nal defens ive war is not signi fi cant . The s i gnificant correla tion possesses no predictive di re ction , however . Externa l War- Attacking Frequent I nfrequent Frequent 46 21 67 I nternal War I nfre quent 45 42 87 91 63 1 54 cp = . 1 71 x2 = 4 . 489 p < . 05 >-. = 0 :>-.. . -1 5 C r = If I were to generali ze , I could say that t he cumulativ e record shows that violence wi thin and without a cultural unit has occurred together . Sumner , Ardrey , and Otterbein ·woul d not f i nd this generalizat io n to the i r liki ng. As anot he r example of discrepancie s between my results and those of Ot terbein--and further evidence that one cannot tell beforehand how resul ts may be different--I agai n quote Otterbe i n 1 s table i Ltercorrelating po l iti cal system , internal 356 war , a nd ini t iating par y, fo lloi,-ed by results of analogous i ter cor--relations from the ODYSSEY . I have :._ igured lar-11-d.a values for Otterbein ' s table . Uncentralized Centralized Political Systems Political Systems Internal War Internal War Cont i nual or Frequent I nfrequent Cont i nual Infrequent or Frequent Initiating Party : Official 3 6 9 9 2 1 1 Anyone 1 2 2 14 2 1 5 8 23 10 3 1 3 cp = . 54 x2 = 6 . 63 :p = - . 27 x2 = . 97 p = . 01 n . s . (Ct terbein 1968 : 283) Uncentralized Cer:..tralized Polit ical Systems Poli t ical Systeills Internal War I nternal War Continual ~ Continual or Frequent Inrrequent or Frequent Infrequent Decision to Commence War : Offic i al 28 27 55 1 9 22 41 or Ccuncil Ar1yone 1 7 1 2 29 3 3 6 45 39 84 22 25 47 cp :::: - . 0735 x2 = . 4536 c:p = - . 0245 .t..2 = . 0282 n . s . n . s . 357 Thus , one of the major significant findings that Otterbe in reports for his 1968 study does not hold up on a larger sample . He can no longer claim that "it is demonstrated that fraternal interest groups and unauthorized raiding par- ties i nfluence the frequency of i nternal war in uncentralized political systems but not in centralized one s " (196 8 : 277) . This is also a rebuttal to the claim that if a correlat ion is significant on a small sample , it will be even stronger on a larger sample . Some are , some are not . One cannot predict. I have the codings to repli.cate other intercorrela- tions of Otterbein ' s i n his 1968 study . However , on the basi s of my criticism i n Chapte r II of this work , its severe deficiencies i n methodology and i nterpretation do not justify further effort or space here . I t i s a shame that Otterbein's tortured and convoluted i nterpretat ions are wasted upon erroneous inductions. The only replication of Ot terbe i n ' s 1970 re s earch that is in the ODYSSEY i s of Otterbein ' s hypothesis that political centralization l eads to military success (1970 : 97) , for which he finds no support at all . Table 21 presents the associ- ation between the s e attributes that comes out of my study , significant at higher than the . 001 level but without pre- dictive direction . Since I have not been concerned wi th "military sophistication ," I have not tested my reinterpreta- tion of Otterbein's results, i n which I hypothesize that at some point militari sm becomes the i ndependent variable . 358 I ndirect ly related to Otterbein ' s 1970 results are mine on the relationship of high casualties to other attri - butes . The following associations are s ignificant at the . 05 level or less . Wr:ere rnili tarisrn i s highly valued , l arge mun- bers of people di e . Where offensive war is freque nt, people die . Where the state exists, people die . Where there are soc i a l classes , fo rmal military organiza tions , authoritative command , or high mili tary expectations , people die . Their deaths may not be c ontri buting to the achievement of mi litary success, however . So casualties i s an i mportant variable but i t is t i ed to a l arger matri x of other vari able s than s i mply military sophi st ication . There i s another area of i nvestigat i on to look at among the wor l d- wi de intercorrelations : t he relationships , i f any , between subsi stence variabl es and warfare variables . Otterbein di smi sses the s i gnificance of any such relationship after pre liminary i nvestigation . I wished to e xamine the re l ationshi p between terrain , mode of subsistence , and type of so ciopolitical system , treated as i ndependent vari abl es , and military organi- zation , tacti cs , and goals of war , treated as dependent variables . . .. As the anal ys i s of the data proceeded , it became clear that ecological and economic factors had little influence , j_n comparison with the type of soci opoli ti cal system, upon the warfare vari ables (Otterbein 1970 :vi i - viii) . Subsequently , Otterbein eliminated discussion of results of these variables from his monograph , rest i ng his generaliza- tions upon the i ndependent variable of soci opolitica l 359 organization a lone . w:ten Russell ( 1972) performed factor anal ysis upon a l arge number of var i able s , he found that a soci al complexity or l evel of technological development factor which l oaded with variables such as subsistence , co□- munity size , socia l stratification , and polit ical autonomy " ... i s only mini ma l ly re l ated to warfare " (1 972: 300 ). A second factor of "formal" variables consisting of descent and re s idence " ... is not at all related t o warfare " (1 972 :300 ). He then goes on to discuss the "psychocultura l" factor that he fi nds most strongl y related to warf are , concluding that formal factors and psychocultural fa ctors pattern independ- ent ly of each other . I, however , have found subsistence to be s i gni f icantly correlated with selected warfare variables , although the lambda values are l arge l y absent, as I have already shown i n i so lat i ng the food-collecting communi ty confi guration . Here I shall go into more detail with regard to t he s ignificant assoc i a tions of agriculturalists . Pastora.lists alone i n the worl d- wi de correlat ions do not associ ate signifi cant l y with a ny other vari able ; they are included i n the l argest class , "food- producers ," i n Table 11. Seeking associations of spe - cific dominant subsistence types , I have correlated them only wi th the 3 types of warfare , specifi c military expectations , and expectations l umped into classes " sub jugation and trib- ute ," "land and plunder ," "trophie s and honors ," and " revenge , defense , or aggressive defense ." Thus, the use of 360 subsistence attribQtes i s not exhaustive , although i i is extens ive . Despite these limita tions , and the weakness of l ambda i n the world- wide run , the associations are i nter- esting . For instance , food- producers taken a ll together do engage i n frequent offensive external war (~ = .1 56 , < .05), and have high military expectations ( cp = . 234 , < . 01); age- sets , military societies , or standi ng armies (cp = . 298 , < . 001 ) ; official de cis ion- maki ng when going to war ( cp = . 260 , < . • 01 ); aut horitat i ve command ( cp = . 395, < .001 ); rewards for warriors(~= . 271, < . 01) ; and a hi gh value on war and violence (cp = .1 66 , < . 05 ). When pas toralists are excluded , however , food- producers are no longer s i gnificant ly associated with any type of warfare . The pastoralists , famous for their external ra i di ng , load the attri but e of food- producers to the point of s i gni ficance. But foo d- producers without pastoralists consisting of i ncipient , extens i ve , and intens i ve agriculturalists , i f i nvo l ved in war , are signi f icantly associated with the military expecta- tion of subjugation ( er.· = . 305 , < . 001 ; Table 27) and the expectation class of subjugation and tribute (cp = . 266 1 < • 001 ) , and are signi f icantly not associated wi th revenge (cp = -. 204) . None of t he relationships of the "agri cultural summary " are predictable . Incipient agriculturalists , such as the Yanomamo , by themse l ves generate a signifi cant nega- tive relationship only with plunder ( cp = -.1 78 ). Exte ns i ve 361 · TABLE 27. INTERCOR..tillLATION OF AGRICULTURE AND THE MILITARY EXPECTATION OF SUBJUGATION Incipient, Extensive, or Intensive Agriculture Present Absent 4 Africa s:: 13 Circurn-Mediterranean 1 Circum-Mediterranean 0 ·rl ~ 8 East Eurasia 1 East Eurasia +' (I) C\l 5 Insular Pacific Cl) bO Q) .:,:J, H 5 South America p.. .g Cl) ~ 0 35 2 s:: 0 ·r-l +' '1l +' t) 0) 18 Africa 4 Africa p. >< 7 Circtun-Mediterranean 6 Circum-Mediterranean til 15 East Eurasia 10 East Eurasia >, +' S--i s:: 19 Insular Pacific 5 Insular Pacific cu (I) +' Cl) 10 North America 19 North America •r-l ,-I ~ 14 South America 13 South America •r-f ~ 83 57 177 Q = .846 ,5 = .305 2 X = 16.419 362 agriculturalists , such as the Tiv , have more associations : with i nternal war(~= . 208 , < . 01) predic table 14 . 5 per cent of the time ; with the specific expectation of trophi es and honors(~= . 223, < . 01 ) , and negatively with the expectation class of revenge, defense , and aggressive defense (~ = - .1 48 , < . 05). I ntensive agri culturalists , such as the J apanese , Hidatsa , and Aztec, by themselves are signi f icantly associ- ated wi th offens ive external war ( ~= .1 53 , < .05). They do not go to war for trophi es and honors(~= -.1 51, < . 05) or for revenge ( ~= -. 306 , < . 001). They do go to wa r for the class of expectations consisting of subjugation and ·tribute (~ = . 377 , < ~001 ). Thi s relationshi p possesses a moderate predictive direct ion : from knowledge that a cultural unit ' s military expectation includes subjugation or tribute , 16 . 4 per cent of the time one can predict t he subsistence mode wi ll be i ntensive agriculture . When one drops out tribute, the asso ciation becomes stronger (~ ~ . 392 , < . 001), as does t he predi ctive direction : 20 . 8 per cent of the t i me , one can predict intensive agriculture from the expectation of subjugation, but not the other way around . I f i nd the direc- tion of the lambdas especially interesting . Without them , we would not know that the expectation of subjugation i s a suf- ficie nt but not necessary condition of i ntensive agriculture . Therefore, there must be unkno1·m i nterveni ng vari ables con- necting subsistence to such a military expectation . The technoenvironmental attribute by it self is not enough to 363 bring out a part icular military circumstance . I n Table 28 , I have i dentif ied the societies . The geographical breakdown by cell i s as follows : a b C d Africa 2 6 2 16 Circum- Mediterranean 1 2 7 2 6 East Eurasia 7 9 2 16 Insular Pacific 2 3 3 21 North Ameri ca 0 3 0 26 South America 1 1 4 26 Cell frequencies 24 29 1 3 11 1 Subj ugation as a specific military expe ctation is a phe nomenon of the Ci rcum- Mediterranean region , the cradle of civi liza- +u l· On . The l ast variable to be dis cussed is probably the mo s t germane demographic one . The i mpo rtance of population den- sity as a variable i n warf are r esearch i s obvious . Prior to the publicat i on of Cross- Cultura l Code 3 (Murdock and Wilson 1972) , I chd not have a source for such i nformation , nor di d I include the vari able as yet 1 more to attempt to co de myself. I have some preliminary results intercorrelating population dens i ty with i nternal war and external war- a ttacldng on the whole Standard Sample . The vari able was not i ncluded i n the or i ginal compute r runs , and I have not yet carried out supplemental runs on both the worl d- wi de and r egional sample s . The followi ng results are based upon c om- putat ions by hand . The attribute s of the variabl e a re : 36 4 TABLE 28. INTE..l'{CORRELATION OF Il'l'TENSIVE AGRICULTURE AND THE MILITARY EXPECTATION OF SUBJUGATION Subjugation Present Subjugation Absent Lozi Turks Burmese Kikuyu Riffians Manchu Nyakyusa Romans Khmer Ganda Gheg Koreans Hausa Basques Tanala Barnbara Armenians Kimam Kanuri Russians Chinese Tallensi Santal Ajie Amhara Kurd Japanese Ot ero Nuba Burusho Ifugao Egyptians Punjabi Javanese Ingassana Lalo Hidatsa Hebrews U. Pradish Balinese Songhai Lepcha Zuni Babylonia Iri sh Inca Fur Vietnamese Aztec Konso Siamese Mapuche 24 Teda Negri. Sero. 29 Nama Chukchee Kutenai Fon Kung Iban G. Ventre Azande Thonga Badjau Pawnee Wolof Mbundu Toradja Omaha Abkhaz Kongo Alorese Huron Basseri Ila Tiwi Creek Palaung Luguru Aranda Natchez Samoans Mbuti Orokaiva Comanche Marshallese Nkundo Kapauku Apache Palauans Banen Kwoma Havasupai Bribri Tiv Manus Papago Cu.T1a Ibo New Ireland Y. Maya Haitians Ashanti Tr obriand Miskito ~ Saramacca Mende Siuai Goajiro Q.) Cl) Shill uk Tikopia Callinago ..0 ·rl Gond Aleut Aymara C/J ~ Toda Montagnais Siriono (!J +' Kazak Micmac Nambicuara ~ H Khal ka Slav Trumai Garo Kaska Timbira Serna Naga Eyak Tupinamba Rhade Haida Botocudo Semai Bellacoola Shavante Nicobarese Twana Aweikoma Andamanese Yurek Cayua Vedda Pomo Lengua Ainu Yokuts Abipon 13 Gilyak Paiute Tehuelche Y11k,,.o-h ·i r- K1 ::.m;:i+.h Yah12:an 111 A r = .208 365 Less than one person per five square miles . Fr om one person per square mile to one per five mi.les . From 1 . 1 to 5 persons per square mile . Fr om 5 . 1 t o 25 persons per square mile . From 26 to 100 persons per square mile . From 101 to 500 persons per square mile . Over 500 pe rsons per square mile . (Murdock and Wilson 1972 : 258 ) Coding is on pop"L;,l a tion density in t he "area exploit ed or contro l Jed by the f ocal or typical community " (Murdock and Wi ls on 1972 : 257- 58 ). I n looki ng at correlat i ons with i nt ernal warfare , one sees a vague c ontinuum with clear extremes . Societies with fe wer t han 1 person per 5 square miles experience little or no i nternal warfare (~ = -. 2128 , p < . 01) and the absence of such war is predictable from population density 17 . 3 per cent of the time . Societies having over 500 pe r sons per square ~ile do experi ence continual or frequent i nternal war ( ~ = . 359 , p < . 00 1) and t hat i s predictabl e 14 .7 per cent of t he time . The i ndivi dual a ttributes of popula tion density between these 2 extremes do not associate signi ficantly wit h i nternal warfare . Associations do devel op , however , when attributes are grouped i nto l arger classes . De nsi t y of 100 persons per square mile appears to be some sort of watershed or boundary . I have di chotomi zed the types of density 2 ways with the followi ng results . First , 366 Populat ion Dens i ty Less than 1 person/5 26 to over 500 sq . mi . to 25/1 sq . mi . persons/sq . mi . Internal War : Frequent 38 43 I nfrequent 51 24 1 56 Q = -. 4-126 ~ = - . 2128 x2 = 1 . 0668 p < . 01 A = . 0746 A C r = . 1733 That is , dens ity of 25 persons or fewer per square mi le is negat i ve ly associated with frequent internal war , making densities of 26 or more pers ons per square mi l e positively associated . The presence or absence of internal war is pre- dictable from knowle dge of dens i ty 17 per cent of the time , and density is predictable from frequency of i nternal war 7 . 5 per cent of the time. Upon dichotomizing another way , how- ever , associ at ion, signifi cance , and predictability i ncrease . Population Density 100 persons or fewer/ 101 persons or square mile more7square rr:ile I nternal War : Frequent 51 30 Inf requent 67 8 1 56 Q = - . 6625 ~ = - . 3069 x2 = 14 . 6952 p < • 001 A = 0 Ar = . 2133 C 367 Now frequent internal war and density of 100 persons or fewer are negat i vely associated , and frequent internal war and density of 101 persons or more are po s itively associ ated at the . 001 level of signi ficance . The presence or absen ce of i nternal war is predictable f rom knowledge of dens i ty 21 per cen t of the t i me and it is not revers ible . Table 29a shows t he resul ts of this i ntercorrelation with geographical dis- tribut i on . The only striking charac terist ic i n such distri- bution that I can see i s t he loadi ng of North American soci- e ties into the Q cell , and t he relatively l arge numbe r of I n~ular Pacif ic societies in t he g cell. Re sults are more limited when intercorrelating popula- tion density wi th offens i ve exte rnal war . Out of all the i ndividual attri but es of populat ion and their possible logi- cal combi nat ions , only 1 attribute--5 . 1 to 25 pe r s ons per s quare mile- -correlates s i gnif i cantly with offens i ve external war , but then only at the . 05 leve l and without predictabil- i ty . Table 29b presents this i nte r correlation , with each cell broken down i nt o it s regional cons tituent s . I cannot di sce rn any part icular geographica l patterning , except for the absence of Eurasian societie s i n the common presence cell . For obvi ous ecol ogica l reas o~s , popul at ion de nsity must be test ed further , e specially within each geographica l region ~ before reaching any conclus ions . TABLE 29a. INTERCORRELATION OF POPULATION TABLE 29b. INTERCORRELATION OF POPULATIOI DENSITY AND INTrillNAL WAl:1. DENSITY AND EXTERNAL WAR- A'I'l'ACKING 100 Persons Or Fewer 101 Persons Or More 5 . 1 to 25 Persons Per All Other Density Per Square Mile Per Square Mile Square Mile Attributes +s::' (!) ::s O' Q) 9 Africa 3 Africa 5 Africa 11 Africa f..., 6 Circum-Mediterr anean 7 Circum-Mediterranean 3 Circum-Mediterranean 14 Circum-Mediterranean ~ rl 8 East Eurasia 8 East Eurasia 16 East Eurasia m 12 Insular Pacific 9 Insular Pacific 2 Insular Pacific 5 Insular Pacific ~ ·rl 4 North America 3 North America 14 North America +' i:: 12 South America 3 South America 3 South America 14 South America 0 u ~ Cl) :;: 51 30 16 74 rl cu i:: H Q) - .p s:: H 6 Africa 3 Africa 5 Africa H (l) 10 Circu~-Mediterranean 2 Circum-Mediterranean 3 Circurn-Mediterranean > Q) 9 East Eurasia 3 East Eurasia 1 East Eurasia 14 East Eurasia $ 5 Insular Pacific 2 Insular Pacific 13 Insular Pacific i::: 127 North America 1 North America 13 North America (l) ::s 10 South Ameri ca 11 South America O' Q) H ~ H 67 8 4 59 Q. = -.6625 /J = -.3069 2 X = 14.695 P< .001 Q. = .5226 /> = .1669 2 X = 4.2687 p~.05 r = .2133 C = .171 C = .297 369 Thi s concludes my analysis of correlations based upon t he whole world and relat i ng my results to those of Naroll and Otterbein where poss i ble . Even i n looking at t he world , I have consistently avoided interpreting the stati s tica l statements as universal generaliza tions but i Lstead have t ried to show where particular geog r aphical regions and therefore particular ecologica l , historical, and t radi t i onal circumstances have produce d the cultural cases that raise a corre l at ion to statistical signi fi cance. I n the next se c tion I shall look closely a t the par ticula r regions themselves . Since to my knowledge no one has examined such variations in print , I will not b e relating my materi a l to that of anyone else . Regional Configurat i ons Before di s cussing the re sul ts of correlat i ons within each g eographica l regio n in detail, a s an illus tration of varying association between the same variables from one geo- graphica l region to another , Tabl e 30 i ncl udes a few excerpts from the exhaustive lis t of phi comparisons i n Appendix D. 370 TABLE 30 . EXCERPTS OF REGIONAL COI-1PARISONS OF PHI FROM APPENDIX D World Africa Ci r-Med Eurasi a I nsular North South Pacific America .iw1erica External war- attacki ng wi th external war-being attackec. : . 500 . 447 . 177 . 354 . 791 . 804 . 378 External war- attacking with military success ( expans i on) : . 402 . 293 . 433 . 410 . 279 .1 80 . 538 External war- attacking wi th high pre st i ge for warrio r : . 306 . 630 . 040 . 280 - . 191 . 461 . 422 Patrilinea l des cent with i nternal war : . 054 . 094 . 1 01 -. 201 -. 468 - . 194 . 514 By i nspe ction of thi s table, and especially the Appendix , one can discern the wide variation i n magnitude and pos itive or negat ive direction of t he two- way association that phi meas- ures . Because of t he l arger numbe r of cases , the phi coeffi- cient s i n the Wo r l d can be l ow and still be s i gnifi cant at the . 05 l eve l . As a r ough gui de , World phi s of . 150- . 155 are significant at this level, while within the regions , a phi must be ab out . 400 to reach such statistical signific ance . Following the procedure used by Driver and Schuessler ( 1967), to obtain a measure of overall s i mi larity and dissim- ilarity in the occurrence of the phi val ues , I have compared the ge ographical regions 2 at a time and measured their joint 37 1 frequencies with Pearson ' s £· I n Tabl e 31, I have f i rst pre- sented Driver and Schue ss l er ' s results, and then my own . The l at ter overall are lowe r i n s i milarity than Driver and Schuess l er ' s , a lthough the co rre l ations between Africa and the Circum- Mediterranean and Africa and the Insular Pacifi c are virtually i dentical i n their magnitude of joint fre- quency . The reader mus t be cautioned , however , tha t su ch measures of compari s on do not measure spe cific vari ables but onl y the di s t r ibut ion of phi va l ues without r egard to the a ttri butes t hey are measuring . Thus the utility of s uch i ntercorrelations of i ntercorre l at ions is s everely limited . I ts virtue here lies primarily, I think , because I can make a f urther comparison to Dri ver and Schuessler' s results . It i s poss i ble that t he warfare variabl es are pulling the degree of similari t y between the regions down , although tha t i s simply a guess . 1 . Africa . The analysis of Africa is based l arge ly upon 19th cen- tury condit i ons , since this i s the period of the first good ethnographic a ccounts . But sub- Saharan Afri ca had already experienced Europe , first t hrough its technology . The flow of guns i nto the region , accelerating t hroughout the century until coloni zat ion , apparent ly radically changed older pat- t erns of conflict, maki ng existing ones more letha l and dif- fus ing militancy throughout t he continent . There i s much 372 TABLE 3-;I • COMPARISON OF DRIVER AND SCHUESSLER I S IrTTER- CORREL/iTI ONS (PEARSON' S £ ) OF THE SIX GEOGRAPHI CAL AREAS WITH THOSE OBTAI NED FROM TI-IE STANDARD SAMPLE 2 3 4 6 5 2 Ci r cum- Mediterranean . 24 . 41 21 . 1 7 . 1 6 Africa . 24 . 57 . 55 . 40 . 46 3 East Eurasi a . 41 . 57 . 56 . 48 . 38 4 Insul ar Paci fi c . 21 . 55 . 56 . 48 . 46 6 South Ameri ca . 17 . 40 . 48 . 48 . 48 5 North Ameri ca . 1 6 . 46 . 38 . 46 . 48 Total s 1. 19 2 . 22 2 .40 2 . 26 2 . 01 1. 94 (Dri ver and Schuessler 19 67 : 341 ) 2 1 3 4 6 5 2 CircuJn- Mediterr anean . 2297 . 2975 . 11 4 3 . 1738 . 17 1 3 Afri ca . 2297 . 2548 . 5556 . 3024 . 2856 3 East Eurasi a . 2975 . 2548 . 2079 . 2386 . 2293 4 I nsular Pacific . 114 3 . 5556 . 2079 . 1 583 . 2477 6 South Ameri ca . 1738 . 3024 . 2386 . 1 583 -3330 5 North America . 17 13 . 2856 . 1193 . 2477 . 3330 Total s . 9866 1. 6281 1 . 2 281 1 . 2838 1 . 2061 1 . 2669 373 material in pri nt within the past 5 years on the hi storiogra- phy of European armaments in Africa . Thi s finely detailed wor k , which has also meant revis io n of earlier and , by com- parison , s i mplistic descriptions of what l ife i n 19th century Africa was about , makes my analysis here also simplist ic . The statistica l data summarize phenomena as they existed i n a geographical regi on a lready exposed to t he diffusi on of par- ticular elements of war . How t hese e l ements were i ntegrated i nto aboriginal culture s was , of course , not nece ssarily dif- fused also . Guns i n most cases were i nco rporated. i nto an older tradit ion . The militant Sudanic states, whose warfare patterns are striking and f amous , are clas sed here as be long- 'i ng to the Circum-Mediterranean region . The 27 Afri can soci eties wi th t he ir Standard Sampl e i dent i ty numbers are : 1 Nama Hottentot 10 Luguru 1 9 Ashant i 2 Kung Bushmen 1 1 Kikuyu 20 Mende 3 Thonga 1 2 Ganda 22 Bambara 4 Lozi 1 3 Mbut i Pygmie s 23 Tallensi 5 Mbundu 14 Nkundo Mongo 28 Azande 6 Kongo 1 5 Banen 30 0toro Nuba 7 Ila 16 Tiv 31 Shilluk 8 Nyakyusa 1 7 I bo 32 I ngassana 9 Hadza 18 Fon 34 Masai Location of each society is i ndi cated by identity number on Map 1 . Africa produces the l argest number of significant i nternal correlations of the 6 regions . Cont i nuing t o use graphic presentation , I have diagrammed the r e l at ionships among the variables within Africa according to t he stronges t l ambda values in Figures 3 and 4 . Where the lambdas are 374 Map 1: Africa •41 4•o 25 • ~~ 29• 30• 31 • .32 28 •33 • •35 •1 3 0 1 • \ 9• • 8 .1 5 • • 6 4 ~ • O 200 600 4oo Boo 0° (adapted from Murdock and White 1.969: 342) This map shows 27 societies classed as African, 16 as Circum- Mediterranean, and 1 (Madagascar) as Eurasian. 375 Fig. 3 Africa: direction and strength of lambdas, focused on 2-4 levels of polity (13). 376 two- w-1y , I have i ncluded the strongest direction i n the fig- ure and both va l ues i n Table 33 . I n Figure 3 , poli t i cal cen- tra lization , indicated by the presence of from 2 to 4 leve l s of jurisdicti onal hierarchy beyond the local communi ty (vari- able 13) , is the a t tribute of pri mary fo cus , whi le i n Fi gure 4 it i s the presence of f rom 1 to 4 such l eve l s ( 14 ) . I n Fi gure 3 there is no re lat ionship between greater polit ica l centrali~at i on and warfar e of any type , but 4 1 . 7 pe r cent of the t i me chi efdoms and states will have some s ort of profes- sional mi l i tary organizat i on ( 24) that is strongl y predictive of a high va l ue he l d fo r warfare (37 ) , and 50 per cent of t he time one can predict that these poli t i es will be e xpandi ng t erritorially (38 ) . The presence of authoritative military command (32) i s better predicted by the pre sence of class strat i f i cat i on ( 19 ) or by complex set tlement patte rn ( 8 ) than by poli t ica l organi zat i on . Conclus ion of war t hrough nego t iat i on i s a bsent ( 27) , predi ctable f rom complex se ttle- ment , and peace ceremonies are absent (28 ) , predict able f rom polit i cal organi zation . Not i ce t hat while t he re l at i onshi p between po lygyny ( 2 ) and polity is s i gnif'i cant , only tau gi ves a predi ct i ve value , and s i nce i t is i n both di re ctions I cons i der it to be tenuous . There i s a l s o a s i gnificant but not a predictive relat ionship between poli ty and the presence of high prestige fo r warriors (33 ) (Tabl e 32) . I have i ncluded the subsidiary configuration of i nternal war ( 21 ) i n this Figure for readability . In t he Afri can case , where 377 local headmen are not hereditary ( 20 ), one - third of the time one can predict frequent i nternal war , and from that , that leaders will be informal (32) and the expectation will be revenge (45) . Those nonhereditary headmen do not conclude their internal wars t hrough negotiat ion (27) , either . Figure 4 i s torturously elaborate , although I hope not snarled. Food-product ion (1 5) , polygyny ( 2) , 1-4 jur i sdic- tional l evels (1 4 ), frequent offensive war ( 22 ), elaborate rewards for warriors (35) , and plunder (43 ) are a ll strong predictors of several other variables and i n some cases of each other . Whereas external war- attacking (22 ) is not sig- nificant wi th "Africa, " within Africa it is one of t he strongest variables . High prestige for warriors ( 33) and high value for war ( 37) appear to be t he depe ndent variables _ost often . There is a direct re lat ionshi p between polity and frequent offens ive war (Table 34 ) as in Figure 2 , but i n Africa the presence of po l ygyny is a bet t er predictor of poli ty t han is frequent war , and patrilinea lity (1 6 ) is the best predictor of polygyny , although food- production i s a close second . Frequent offensi ve war is c ompletely predict - able from elaborate rewards for warriors (35 ), predictable half the time from patri l ineality (16) , half the time fr om plunder as a military expectation (43 ) as i s defensive war- fare (23) , and predictable one-third of the time from patri- locality (3) . In turn , both plunder (43) and a high value for war (37 ) are strong predictors of high prestige (33 ). 378 Fig. 4 Africa: direction and strength of lambdas, focused on 1-4 levels of polity (14). 379 Both rewards and pl under , and expectations of violence (36) and a high value for war are mutually predictable at least 50 per cent of the time , althougn trophies and honors ( 44) is the strongest predictor of the expectation that vio l ence i s a solution to problems (36) . The strongest predicto r of a high val ue for war remains the presence of professional military organization- - age- sets , military societ i es , and standing armies ( 24 ) . Not i nclude d in the diagrams are the find ings that communities of l ess than 100 persons are predictably food-collectors 50 per cent of the time , and that where there is any type of warfare , communit ies of s uch s ize engage in it i nfrequently if at a ll. This i s the only demographic at tri- bute that has significant ass ociations in Africa . Warfare in Africa appears to be characteri stically offensive , result i ng in elaborate rewards for individual warriors who have fought usually for pl under . Subsequent ly, there are dependent variables of prestige, expectations of violence as a solution to problems , and war as highly valued . Most of t he lambdas in Africa are two- way : there are cons i s - tent feedback relations amo ng the vari ables (Table 33) and several are exactly mutually predictable . Thi s complicates the confi guration in Figure 4 and i s eviden ce against any simple one-directional sequence of relationships . That is , there is circular causali ty at several points. 380 TABLE 32 . SIGNIFICANT CORRELAT IONS WITH O LAMBDAS--AFRICA Pa i rs of Variables cp p P.. o~r attribute Pol ygyny ( 2) with : 2- 4 l evels beyond the local community (1 3) . 418 . 05 Mili tary success (territoria l expansion)(38 ) . 454 . 02 Trophies and honors (44 ) . 399 . 05 Ro~ attribute 2- 4 Levels Beyond the Loca l Community (13) with : High pr estige for warriors (33) . 447 .05 Row attribute Frequent External War-Att a cking (22) wi t h : Trophie s and honors (44 ) . 408 . 05 381 1'ABLE 33 . SIGNIFICANT CORRELAT IONS WITH LAJVIDDAS I N TWO DI RECTI01·S- -AFRI CA Pairs of Variables cp p Ar A C Row attribute Polygyny (2) with : Food- producers (15) . 705 . 001 . 571 . 250 Patrilineality (1 6 ) . 792 . 001 . 667 . 500 Row attribute Complex Settlements (8 ) with : 2-4 l evels beyond the community(13) , 434 . 05 . 300 • 22.2 Class stratification (1 9) . 613 . 01 . 500 . 545 Conclusion by negot i ation ( 27 ) -.472 . 05 . 375 . 444 Authoritarian command ( 32 ) . 548 . 02 • 500 .545 Row attri bute 2-4 Levels Beyond the Local Community (13) with : Class stratification (19) . 583 . 01 . 333 . 455 Age- sets ... standing armies ( 24 ) . 462 . 01 . 222 . 4 17 Peace ceremony ( 28) -. 816 . 01 . 750 . 667 Authoritari an command (32) . 462 . 05 . 333 . 455 Military success ( expansion)(38) . 597 . 01 . 444 .500 Row attribute 1-4 Levels Beyond the Local Community (14) with : Frequent external war- attacking . 4 76 . 02 . 286 . 167 (22) Prest i ge for warriors ( 33 ) • 55 3 • 02 . 200 . 429 Rewards for warriors (35) • 564 . 02 . 400 . 250 Expectations of violence (36) . 592 . 01 . 167 . 400 Plunder (43) . 584 . 01 . 429 . 200 Row attribute Class Stratification (19) with : Military success (expansion)(38 ) . 457 . 02 . 364 . 300 Row attribute Hereditary Local Head- man (20) with : I nternal war (21) -. 415 . 05 . 200 . 333 Authoritarian command ( 32 ) . 462 .05 . 333 . 455 Row attribute Internal War (21) with : Authoritarian command ( 32 ) -. 462 . 05 . 333 . 445 -Revenge ( 4 5) . 463 . 02 . 417 . 462 382 TABLE 33 ( Continued ) Pairs of Vari able s cp p Ar A C Row attribute External War- At t a cki ng ( 22 ) with : External war--be i ng attacked (23 ) . 44 7 . 05 . 167 . 286 Prest i ge f or warriors ( 33 ) . 630 . 01 . 400 . 500 Expectat i ons of violence (36 ) . 632 . 01 . 333 . 500 High va l ue for war ( 37 ) . 517 . 02 • 1 67 . 444 Plumier (43 ) . 65 2 . 01 . 500 . 400 Row att ribute Frequent Externa l War- Being At tacked (2 3 ) with : Pl under (43) . 588 . 01 . 429 .200 Row at t r i bute Age- Sets ... Armi es (24) wi t h : High prestige for warri ors ( 33 ) . 539 . 02 . 500 . 286 High value fo r war ( 37 ) . 840 .001 . 818 . 800 Rov.r attribute Conclusion by Negot i- ation (2 7) wi th : High prest i ge f or warri ors ( 33 ) - . 620 . 01 . 556 . 200 Hi gh val ue for war ( 37) - . 4 71 .05 . 444 • 1 67 Military succe ss ( expansion ) ( 38 ) -. 472 . 05 . 444 . 375 Land (42 ) -.472 . 05 . 444 . 375 Row attribute Military Expe ctat ions I (29 ) with : Expe ctat i ons of violence ( 36 ) . 596 . 01 . 500 • 556 Row attr ibute Military Expe ctations II ( 30) wi th : Military suc cess ( expansion) ( 38 ) . 592 . 01 . 545 . 500 Land (42 ) . 848 . 001 . 8 19 . 778 Row attribute Hi gh Casual ties (31) with : Tribute (41 ) . 784 . 01 . 667 . 500 Rmv at t r i but e High Prest i ge f or Warri ors ( 33) with : Elaborate rewar ds for warri ors (35) . 564 . 02 . 400 . 250 Pl under (43 ) . 780 . 001 . 71 4 . 600 383 TABLE 33 (Continued ) Pairs of Variabl es cp p "- r "- C Row attribute Elaborate Rewards for Warriors ( 35 ) with : High value for war (37) . 555 . 05 • 250 . 400 Row attribute Expectations of Violence (36) with: Trophies and honors (44 ) . 724 • 01 . 667 . 500 Row attribute Military Success (Expansion) ( 38 ) with : Land (42) . 422 . 05 . 300 . 222 Trophies and honors (44) . 422 . 05 . 300 . 222 384 TABLE 34. AFRICA: INTERCORRELATION OF LEVELS OF JURISDICTIONAL HIERARCHY BEYOND THE LOCAL COMMUNITY AND EXTERNAL WAR-ATTACKING Continual/Frequent External War Infrequent External War Nama Hottentot Kongo Thonga Nyakyusa Lozi Mbundu Ila Ganda Tiv Fon Mende Ashanti ti) Bambara rl (l) Azande :> (1) Otoro Nuba H Shilluk Masai 15 2 Ki kuyu Kung Banen Luguru Ingassana Mbuti Pygmies Tallensi 3 4 Q = .818 p = .476 x2 = 5.445 p = .02 C = .43() :\ = .286 C 385 2 . Circum-Mediterranean . The configurations in this region appear to be less complex than those in Africa ; there are fewer significant associations and generally the lambda va lues are lower . This is in c ontrast to the large number of variables in the world- wide i ntercorre l ations that s i gnificantly associate with 11 Ci rcum- Mediterranean . 11 The representative societies i n this region are spread more widely through time than those i n Africa . Anc ient Medi terranean states , Sudanic states , and modern states are i ncluded . The 29 societies in the region are : 21 Wolof 38 Bisharin 47 Turks 24 Songha i 39 Nubians 48 Gheg Albanians 25 Fulani 40 Teda 49 Romans 26 I-iausa 4 1 Tuareg 50 Basques 27 Kanur i 42 Riffians 51 I rish 29 Fur 43 Egyptians 52 Lapps 33 Kafa (1200 BC) 54 Russians 35 Konso 44 Hebrews 55 Abkhaz 36 Son::.ali 45 Babylonians 56 Armenians 37 Amhara 46 Rwa l a Bedouins 5? Kurd Loe at iorJ. of' each society is given by i dentity number on Maps 1 and 2 . The Kafa do not turn up in any of the corre- lations with the warfare variables because of unavailabi lity of data and the absence of an alternate society . In Figure 5 , 2- 4 levels of jurisdictional hierarchy (13) i s not the powerful predictor of other variable s as in Africa . Instead , the attributes of violence as an expect ed solution to problems ( 36) and the presence of a high va lue for war (3 7 ) are predictors of the state , while the state is 386 Map 2 . -- Circtnn-Medi terranean* ' \ \ \ ' \ \ 6'\ 5 •' ' \ \ \ \ ) o 400 Boo 200 6oO Miles 200E (adapted from Murdock This map shows 13 societies classe as Circurn-Medi terranean a.rid 5 as East Eurasian. · ifMurdock and White identify this map as "West Eurasia" and Map 3 as "East Eurasia," apparently because it is not practical to locate all and only those -societies of a particular geographic region on 1 partial map. 387 Fig. 5 Circum-Mediterranean: direction and strength of lambdas, focused on 2-4 levels of polity (13). 388 t he predictor of the presence of class stratification (19 ) and prof eosional military organiza tio n (24 ). Authorita t i ve command (32 ) is best predicted by the presence of subj ugation of peopl e and territory a s a military expectation (40) , and that in turn is pre dicted by class s tratification (19). Where t here is authoritative command , there is little or no prestige gained by individual warriors or soldiers . There i s no relat ionship be~ween the state and military success ( 38 ), nor i s there a direct relationshi p between the state and fre - quent offens ive war (22). There i s , however , a significant but non- pr edictive relationship between frequent offens ive i'iar and military success ( cp :::: • 4 33 , < • 05) . The Ci rcum- Ne diterranean state does not go to war for tribute (41) or for trophie s and honors (44). I n Figure 6 , societies with 1 to 4 jurisdictiona l leve ls beyond the local community (14) have no s i gnificant relat ionshi ps with settlement type , class stratification , type of military mobilization , command , and dec ision-maki.ng . This class of politie s , however, i s directly associated with and predict i ve ~f fre quent offens ive war (2 2) , where the Sudanic s t ates are having an effe ct (Table 35) . Where in Figure 5 the presen ce of the state is predicted by expecta- tions and value s of violence (36 and 37) , i n Figure 6 the strongest predictive direction is reversed , so that expecta- tions and values are predi.cted by type of polity . The l ambdas her e are i n the same direction but stronge r t han 389 · Fig. 6 Circum-Mediterranean: direction and strength of lambdas, focused on 1-4 levels of polity (14). 390 ✓ those i n the world- wide conf i gurat io n (Fi gure 2 ) . I n this r egion t here i s a ve ry strong predictive relat ionshi p between e xpe ctat i ons and value s of violence , and both are predict ors of high casualties ( 31 ) . Frequent offens i ve war can be pre - dicted equally we ll by either polity or the military expe c- tations of subjugation , tribute , or l and (29) . In striking contras t to the African patter n , where i nt e rnal war i s asso- ciated with the absence of authoritat i ve command and. the pres- ence of nonhereditary loca l headmen , in the Circum- Me diterranean region i nt ernal war (21) i s be s t predicted by the presence of elaborate reward s for fighters (35) , plunder (43 ), a nd polygyny (2). There i s no relationship with res i- dence , des cent , or local political organization . I nstead , personal gain a nd prestige s eem to be t he keys . Circular causality must be i nvolved i n t hese configurations also , although the var iables with predictive values in both direc- tions are not the same as those i n Africa (Table 36 ). Actu- ally , t his i s the case throughout the geographical regions : t he attrlbutes vary , as do the dire ction of the strongest prediction and the predictive value s themse l ves . 391 TABLE 35. CIRCUM-MEDITERRANEA..1'.I: INTERCORRELATION OF LEVELS OF JURISDICTIONAL HIERARCHY BEYOND THE LOCAL COMMUNITY AND EXTERNAL WAR-ATTACKING* Continual/Frequent External War Infrequent External War Wolof Songhai Fulani Turks Hausa Gheg Albanians Kanuri Fur Somali Am.hara Bisharin Teda U) Tuareg rl (J) Riffians :> (J) Egyptians ....:I Hebrews .:::t I Babylonians ,- Rwala Bedouin Irish Russians Kurd 18 3 Nubians Lapps 0) rl Cl) :> (lJ ....:I 0 0 2 = p = • 586 2 Q, 1 • 00 X = 7 • 886 p .( • 01 C = A = • 400 * The Basques and Romans are not included in thii t able because the Ethnographic Atlas has not coded them for jurisdictional levels. 392 T_.BLE 36 . SIGNIFICANT CORRELATIONS WI TH LAMBDAS IN TWO DIRECTIONS--CIRCUM-MEDITERRANEAN Pairs of Variabl es p \ r \ C Row attribute Complex Se t tlements (8 ) with : Authoritative command (32) . 675 . 01 . 625 . 571 Row attribute 2- 4 Levels Beyond the Loca l Community (13) with : Class stratification (1 9) . 610 . 01 . 375 . 545 Age- sets ... armies (24 ) . 564 . 01 . 286 . 500 High expe ctat ions of violence (36 ) . 500 . 01 . 500 . 333 High value for war ( 37) . 468 . 05 . 286 . 1 67 Subjugation of people (40) . 590 . 01 . 143 . 455 Row attribute Unili neality (17) with : Frequent internal war ( 21) . 458 . 05 . 417 . 300 Subj ugation of people (40) -. 564 . 01 . 500 - 538 Row attribute Class Stratification (1 9) wi th : Age - sets ... armi es (24 ) . 578 . 01 . 444 . 500 Subjugation of people (40 ) . 632 . 01 . 500 . 615 Row attri bute Frequent I nternal War ( 21) wi th : High prest i ge for warriors (33 ) . 428 . 05 . 300 . 364 Plunder (43 ) . 598 . 01 . 400 . 250 Row attribute Frequent Exte rnal War-Attacki ng (22 ) with : Mi litary Expectations I ( 29) . 600 . 01 . 400 . 250 Military Expectations II ( 30 ) . 600 . 01 . 400 . 250 Row attribute Age - Sets .. . Armies (24) wi th : Authoritative command (32) . 471 .05 . 286 ♦ 375 Sub·ugation of people (40) . 678 . 001 . 556 . 636 Row attribute Peace Ceremony (28) with : Subjugation of people (40) -. 791 . 02 . 667 . 750 393 TABLE . 36 ( Continued ) Pairs of Variables cp p Ar Ac Row attribute Authoritat i ve Command (32) with : High pre st ige for warr·iors (33) - . 492 . 05 . 375 . 444 Subjugat i on. of people ( 40) . 685 . 01 . 625 . 571 Row attribute High Prest i ge (33) with : Land (42 ) . 484 . 05 . 375 . 286 Rm: attribute El aborate Rewards ( 35) with : Land (42 ) . 463 . 02 . 417 . 462 RoK attribute Expe ctations of Vi olence (36) wi th : High value for war (37) . 886 . 001 . 800 . 833 3 . East Eurasia The 34 societies in this geographical region possess probably the greatest cultural and temporal diversity of the 6 areas. There are 19th and 20th century tribal societ ies ar:d both ancient and 20th century states of various sizes . The representatives. with the~r i dentity numbers , are : 53 Yurak Samoyed 69 Garo 81 Tanala 58 Basseri 70 Serna Naga 82 Negri Sembilan 59 Punjabi 71 Burmeee 114 Chinese 60 Gond 62 Palaung 11 5 Manchu 61 Toda 73 Vietnamese 11 6 Koreans 62 SantaJ. 74 Rhade 11 7 J apanese 63 Uttar Pradesh 75 Khmer 11 7 Ainu 64 Burusho 76 Siamese 11 9 Gilyak 65 Kazak 77 Semai 120 Yukaghir 66 Khalka Mongols 78 Ni cobarese 1 21 Chukche e 67 Lolo 79 Andamanese 68 Lepcha 80 Vedda 394 Several of these societies at the target dates were experien- cing nationali st or imperialist warfare, i n addition to the usual displ acement warfare of the modern colonial period . Information was not cons i dered from warfare directly i nvolving European groups , except i n t he case of the Chukchee , who expanded successfully against the Russians , as well as the Yukaghi r , and are one of the few tribal societ i es to suc- ·ceed against a state . The societies are locat ed by i dentity nmnbers on Maps 2 , 3 , and 4 . The configurat i ons of Eurasian warfare are remarkably different from those of the rest of the worl d be cause there is no di rect relat i onship between any form of politica l organization beyond the local community and any type of war- fare . In Figure 7 , from the presence of offensive war (22) or high prest ige for warriors (33 ) , one can predict expecta- tions and values of violence (36 and 37) . Where there i s mil- itary success (38) or the expectations of land and plunder (50) , one can make a moderate l y strong prediction that fre - quent offensive war will exist (22 ) . Where violence is he l d as an expected solution to problems (36), not quite one- third of the time one can predict that political organi zation will be a chiefdom or state (1 3 ) . Such political centralization i n turn is a good predictor of the presence of class strati- fication (1 9) and a moderate predictor of complex set tlement s (8) and the initiation of war through agreement or announce- ment ( 26 ). The presence of some form of professional 395 Map 3: East Eurasia I I I I • 68 ~ 69 • .70 .12. #79 ~ .18 0 400 ~- I 100°E ·,• . 1200E -200 (adapted from Murdock and White 1969: 344) This map shows 25 societies classeQ as East Eurasian and 1 as Insular Pacific. 396 0 ' Fig. 7 East Eurasia: direction and strength of lambdas, focused on 2-4 levels of polity (13). 397 military organization (24 ), t he absence of pea ce ceremoni es ( 28) , and the presence of authoritat i ve command ( 32 ) are the best pre dic t ors of the presence cf the state . Professional military organization ( 24) or food- production ( 15) i s a better predictor of the presence of complex settl ements (8 ) t han the state (1 3 ). Aut horitative command ( 32 ) is the be s t predict or of professi onal military organiza tion , and f ollows i t as the se cond- best predictor of the presence of t he s tate . There i s l ittle di fference i n pat t e r n between Fi gure s 7 and 8 . When the po l it i cal attribute i s 1 to 4 j uri sdic- tional l eve l s beyond the l ocal communi ty (1 4 ), the only new ciire ct re l at i onshi p i s wi th f ood- product ion (1 5) : where food- production i s the dominant subsistence mode , 45 per cent of the time there wi l l be 1 to 4 l eve l s of polity outs i de t he local community . Other differences between t he 2 Figure s a re i n predictive di re ct i on and strength . I n general , the Eurasi an warfare configurations appea r no t to i nclude the androcentri sm , i ndi vidualism , or material- i sm of the regions considered so far . Moreover , the predict - ability between expectat i ons of viol ence and hi gh val ue fo r war is very low compared to the other regions . The patt er n- i ng of a l ow value for vi olence and war i sol ated when ge ogr a - phy was i ntercorrelated as a variable holds up within the re gion i D spite of a l l those ho rdes out of Asia . I nternal warfare has a tiny l i ttle configurat i on of i ts own and does not tie icto the larger patterns at all . I woul d be 398 ' Fig. 8 East Eurasia: direction and strength of lambdas, focused on 1-4 levels of polity (14). 399 espe cially intere s ted to see if the Eura sian pa ttern t rac~d here holds up on a large r sample . Tabl e 37 g ives the lambda values t hat occur i n both directions . TABLE 37 . SIGNIFICANT CORR.ELATIONS WITH LAMBDAS I N TWO DIRECTIONS--EAST EURASIA Pairs of Variables cp p Ar A C Row attribute Complex Settlements ( 8 ) with : 2- 4 levels beyond l ocal comrr:u- .487 .01 . 400 . 308 nity (1 3 ) 1- 4 l evels beyond l oca l commu- . 516 . 01 . 46 7 . 273 nity (1 4) Clas s strat ification (1 9 ) . 525 . 01 . 500 . 364 Mi litary expectations I ( 29 ) . 340 . 05 . 267 .21 4 Row attri bute 2- 4 Levels Beyond the Local Community (1 3) with : Class stratification (19) . 786 .001 . 625 . 727 Age - sets ... armi es (24) . 651 . 001 . 61 5 . 583 Peace ceremony (28 ) -. 683 . 02 . 600 - 333 Authoritat ive command ( 32) . 614 . 01 . 583 . 444 Expe ctat i ons of violence (36) . 422 . 05 . 300 . 222 Row attribute 1-4 Leve l s Beyond the Loca l Community (14) with : Food- pr oducers ( 15 ) . 577 • 001 . 455 . 143 Class strat i fi cat ion (1 9 ) . 497 . 01 . 273 . 273 Mili tary expectations I (2 9) . 351 . 05 . 091 . 231 Authoritat ive c ommand ( 32 ) . 646 . 01 . 42 9 • 556 Row attribute Class Stratificat i on ( 19) with : Age-sets . .. armie s (24 ) . 578 . 01 . 500 . 444 Authori tat ive command ( 32 ) . 552 . 02 .500 . 375 Row attribute Frequent External War- Attacking ( 22 ) with: Expecta tions of violence (36 ) . 566 . 01 . 538 . 333 High value for war (37) . 632 . 001 . 571 . 250 Land and pl under (50) . 459 . 01 . 438 . 308 400 TABLE 37 (Cont inued ) Pairs of Variables p >-r >- C Row attribute Age - sets ... Armies ( 24) with : Authoritative command ( 32 ) . 721 . 001 . 667 . 556 Subjugation of people (40) . 455 . 02 . 385 . 1 1 1 Row attribute Officia l Decision ( 25 ) with : Authoritat ive command (32 ) . 574 • 01 . 200 . 429 Rm,; autribute High Prestige ( 33 ) with : High value for war (37) . 659 . 001 . 400 . 500 Row attribute Expectations of Violence (36 ) wi th : High v a lue for war ( 37 ) . 458 . 05 . 286 • 167 4 . I nsular Pacific . Most of the 31 societies in this region have as target da tes the bes t descriptions made prior to severe culture change due to European contact i n the 19th and 20th centuries . The temporal and cultural variety is not so dramati c as that within the Eurasi an region . The societies ar~: 401 83 Javanese 94 Kapauku 104 Maori 84 Balinese 95 Kwoma 105 Marque sans 85 Iban 96 Mar:.us 106 Samoans 86 Bad jau 97 New Ireland 107 Gilbertese 87 Toradja 98 Trobri anders 108 Mar shal lese 88 Tobelore se 99 Siuai 109 Trukese 89 Alorese 100 Ti h.opia 110 Yapese 90 Ti wi 1 01 Pe nte cos t 1 1 1 Palauans 91 Aranda 102 Mbau Fijians 11 2 Ifugao 92 0rokaiva 103 Ajie 11 3 Atayal 93 Ki mam They are l ocated by the i r i dent ity numbers on Maps 3 and 4 . I n thi s region politica l organizat ion has re l at ively little to do with warfar e . I n Figure 9 , one can pre dict the pre se::ice of the state i n some form (1 3 ) from t he presence of subjugat i on of peoples (40) as a military expe ctation 50 per ce~t of the t ime . Sub jugation is the i ndependent variable most often i n this confi gurat ion . High casualties ( 31) ca n be predict ed equally well by e i ther the presence of t he state or by military succe ss (territoria l expansi on) ( 38 ) , and mil- i tary success i n turn is predict ed by the spe cific military expectation of l and (42) . Th'e presence of the sta te is a s econdary predictor of authoritarian command ( 32 ). Note that i n this configurat i on vio l ence as an expected so lution to problems and a high value for war ( 36 and 37) are s i gnifi- cantly absent , the former predicted by the presence of con- clusion of war by negotiation ( 28 ) and the l atter by s ome form of professiona l military organization ( 24 ) . I n Figure 10, there i s no longer any predictable asso- ciat ion between att i tudes toward war and attributes of mili- tary and politi ca l organi zat ion . What remains i s an 402 Map 4: Insular Pacific ., . . - . , .., .. .. . () ., 20°s 40°E (adapted from Murdock and White1969: 345) This map shows 2 societies classed as East Eurasian and 28 as Insular Pacific. 403 Fig. 9 Insular Pacific: direction and strength of lambdas, focused on 2-4 levels of polity (13). J 404 Fig. 10 Insular Pacific: direction and strength of lambdas, focused on 1-4 levels of polity (14). 405 association wi th high prestige for i ndividual warriors (33) . Offens i ve war ( 22) i s the independent vari able most often i n predictions about i ts ass ociations , yet there are only 3 attributes that it predicts. Where there i s frequent offen- sive war , communi ty size will be something other than 100-400 persons (11), the subsistence mode will be i ntensi ve agricul- t ure , and there will also be frequent attacks by other cul- t ural unit s (23 ) . Tab l e 38 i s the intercorrelat ion of exter- nal. war - a t tacki ng and i ntens ive agriculture. There are only 2 cases i n the common presence cell, the J avanese and the Ajie . They are , however , the only i ntensive agriculturist s in the i ntercorre lation . Expanding the subsistence variabl e to i nclude all agriculturalists , but not pastorali sts , does not produce a signifi cant correlation. As a whole , the conf i gurations of variable s i n the I nsul ar Pacific are patchy . For i ns t ance , in Figure 10 the only warfare i ntercorrelat ions s i gnificant at the . 01 leve l or higher are between complex se ttlement s (8 ) and high casual - ties ( 31) , 1 to 4 levels of jurisdiction (14) and authorita- tive military command ( 32 ), high prestige fo r warriors (33) and expectations of violence as a sol ution to problems (36) , and offensive war and defens i ve war (22 end 23 ) . Militarism i s relatively weakly deve loped i n this region , and practices and institutions of reco nciliation are present . The opposite is the case i n the Ci rcwn- Medi terranean region . Table 39 shows tho se re l ationshi ps that generate lambdas i n both directions . TABLE 38. 406 INSULAR PACIFIC: INTERCORRELA.TION OF INTENSIVE AGRICULTURE AND EXTERNAL WAR-ATr.A.CKING Continual/ Frequent External War Infrequent External War J avanese Aj ie 2 0 Iban Badjau Orokaiva Alorese Pentecost Tiwi Marque sans Aranda 'rrti."l{ese Kapauku Kwoma Manus Trobrianders Tikopia Maori . Marshallese Yapese Palauans 5 13 2 Q - 1.00 d = .454 x = 4.127 A = .286 C 407 TABLE 39 . SIGNI FICANT CORRELATIONS WITH LAMBDAS I N TWO DIRECTIONS-- I NSULAR PACIFIC Pairs of Variabl es p A C Row attri bute Compl ex Set t l ements (8 ) with : Military expectat i ons III ( 30) . 439 . 02 . 250 , 308 Hi gh casualtie s ( 31) . 65 1 . 01 • 500 . 42 9 Row att r ibute Mean Community Size 100- 400 Persons (11) with : Frequent External Wa r-Atta cki ng ( 22 ) - . 471 . 05 , 375 . 286 Row at tribute 2- 4 Levels BeyoTid the Local Community (1 3 ) with : Hi gh casual t i es (31) . 535 . 02 . 200 .429 Subjugat i on of people (40 ) . 668 . 001 . 500 . 400 Row attribute 1- 4 Leve ls Beyond the Local Community (1 4 ) with : Class stratification (1 9 ) . 665 . 001 , 538 . 455 Row at t ribute Frequent External Wa r-Attacki ng ( 22 ) with : Frequent external war--being attacked (23) . 791 . 001 . 667 .750 Row att ribute Frequent External War - Being Att a cked ( 23 ) with : Military expectations I ( 29) . 440 . 05 . 400 , 333 Row attribute Co nclusion by Negot i- a tion ( 27) with : Expe ctat ions of violen ce ( 36 ) - . 567 . 02 . 500 . 333 Hi gh val ue for war ( 37 ) -. 523 . 02 • 500 • 333 Row attri bute Military Expectat ions I ( 29 ) wit h : Expectati ons of violence ( 36) . 471 . 05 . 375 . 286 408 Tli.BLE 39 (Cont i nued ) Pairs of Variables cp p "-r "- C Row attribute Hi gh Prestige (33) with : Expectations of violence (36) . 599 . 01 _375 . 286 Row at t ribute Expe ctations of Vi ol ence ( 36) with : Hi gh val ue f or war ( 37) . 587 .02 . 429 - 556 5. North America . The 33 so ciet i es i n North America are a l l Ameri can I ndians prior to European contact or absor ption . They are : 122 Ingalik 133 Twana 144 Huron 1 23 Aleut 134 Yurok 14 5 Creek 124 Copper Eski mo 135 E . Pomo 146 Natchez 125 Jviontagnais 136 Yokuts 147 Comanche 1 26 I'fiicmac 137 Paiute 148 Chi r i cahua Apache 127 No . Saul teaux 138 Klamath 149 Zuni 128 Sl ave 139 Kutenai 150 Havasupai 129 Kaska 140 Gros Vent r e 151 Papago 130 Eyak 14 1 Hi datsa 152 Hui chol 131 Haida 142 Pawnee 153 Azte c 132 Bella cool a 143 Omaha 154 Popol uca One coul d only wish that the superior quality of North Ameri can ethnographies were c or~sistent l y matched by materi a l from the rest of the world . Location of societies i s g i ven by identity nwnber on Maps 5 a nd 6 . '.l'he configurations of attributes in North America are more complex than those in East Euras i a and the I nsular Pacific , but not so c omplex as those in Africa and South 409 Map 5: North America 40°N • 140 • 141 • 20°N <::;:::, 0 400 800 4 I I 200 '1oo M 12O0w (adapted from Murdock and White 1969: 346) This map shows all 33 societies classed as North American and 2 as South American. 410 America . There are 10 North American cases that have 1 to 4 l evels of jurisdictional hierarchy beyond t he local CO[[I.IllU- nity, and of these only 2- - the Creek and the Aztec--h2.ve 2 _to 4 levels . One distinctive characteristic of this region is that warfare is predominantly defensive , and the attribute of frequent external war- attacking is usually predicted by other attributes rather than being the predict or , as i s t he case i n t he world- wide , African, South American , and even I nsular Pacific configurations . In Fi gure 11, some form of central- ize d pol i ty (13) , i . e ., the Creek and the Aztec , predi cts t he presence of complex settlements (8), authoritarian military command (32) , and high prestige for warr iors (33 ), whi le the presence of the state is predicted _gy the specific military expectation of tribute (41) (the Aztec at work) . The state- - or proto- state--predicts the presence of military command , milita ry command · predicts the presence of hereditary local headmen (20 ), and hereditary local headmen predi ct frequent offens ive war (2 2) , although the last is most strongly pre- dicted by frequent defensive war (23) , and secondarily pre - dicted by elaborate rewards for warriors and high pre st ige for warriors (3 5 and 33) . The state confi guration a nd the androcentric configurat ion appear to be blending i nto each other . Internal war (21) is significantly associated wit h exogamous community ( 5) (cp == . 545 , < . 01 ) and is predicted by such community organization 33 per cent of the time . Inter- nal war is tied into the larger North American confi gurat ion 411 l'ig. 11 North America: direction and strength of lambdas, focused on 2-4 levels of polity (13). 412 only through complex sett l ements (8) : where there i s fre - quent i nt ernal war , compl ex sett lement s will exist 20 per cent of the time . The l ast po i nt seems nebulous, but recall that i n the world- wide i ntercorre l ations , internal war i s significant l y absent in North America in t he f i rs t pla ce . I n Figure 12 , the presence of 1 to 4 levels of juris- diction beyond the local community (1 4 ) is a predi ctor of fre quent offensive war ( 22) (Table 40) , hi gh prest i ge for warriors (3 3), a high value for war (37 ) , and freque nt defen- sive war ( 23 ), while it i s in turn pre di cted by authori tarian mi litary command (32) . However , the presence of the military e xpe ct2.t ions of subjugat ion , tribute , l and , or trophies and honors (29 ) predict s freque nt offensive ~ar , hi gh pre st ige for warriors , and a h i gh value for war equally we ll. This class of military expe ctations predicts the presence of the specific military expe ctation of trophi es a nd honors (44 ) 75 per cent of the time . That i s , the class of expectations has signi f icant associations at all only be ca use of the inclusion of trophi es and honors i n the class ; without this famous North American tra i t , the class does not re l ate significant l y to anything . Polity (1 4) predicts the pr esence of high pres- tige (33) 36 per cent of the time , a nd hi gh prestige predi cts po s i t ive expectat ions of vi ol ence (36 ) 50 per cent of the time . Along another path , defensive war (2 3) predict s a high value for war nearly 54 per cent of the t i me , and high valu- ation (37) predict s expectaiions of violence (36 ) hal f the 413 0 Fig. 12 North America: direction and strength of lambdas, focused on 1-4 levels of polity (1 4). 4-14 TABLE 40. NORT'.rI AMERICA : INTERCORP..ELATION OF LEVELS OF JURISDICTION.AL HIERARCHY BEYOND THE LOCAL COMMUNITY .AND EXTERNAL WAR-ATI'ACKING Continual/Frequent External War Infrequent External War Micmac Kutenai Eyak Omaha Gros Ventre Par,mee Huron Creek Natchez Aztec Cl) rl OJ :> (l) ...:i ~ .,.I.. 8 2 Aleut Ingalik Haida Copper Eskimo Klamath Montagnais Hidat sa No. Saul teaux Comanche Slave Chiricahua Apache Kaska Zuni Bellacoola Papago Twana Cl) r-1 Yurok OJ :> E. Pomo Q.) ...:l Yokuts 0 Paiute Havasupai Huichol 8 14 Q. = .750 /, 2 = .405 X = 5.236 p< .05 C = .375 "- = .375 C time and frequent offensive war 46 per cent of the time . Defensive war als o predicts frequent offensive war (2 2 ), tro- phies and honors (4 4) , and rewards for warriors (3 5), the core of t he defensive pa ttern . There are 2 unusual intercor- relations that I have iucluded here , even though they are based upon only 11 and 9 cases , respective l y . Matrilineal descent (16) has a s i gni f icar,t assoc i ation with both frequent offensive war and frequent defensive war . I have included the tabl es i n the text . Externa l War- attacking Frequent I nfrequent Des cent Patrilineal 0 3 Matrilineal 7 1 11 Q = -1 . 00 cp = - . 810 x2 = 1.219 p < • 01 \ = . 750 \r - . 667 C External War--Being Attacked Frequent Infrequent Descent Patrilineal 0 2 Matrilineal 6 Q = -1 . 00 cp = -.7 56 x2 = 5 . 143 p < . 05 \ = . 667 \ r = • 500 C 416 On the one hand these results may be spurious , since the Ns a.re so small . On the other hand , matrilineal des cent i s a rare condi tior, and one cannot help but wonder why nearly all matrilineal societies in North knerica are engaged i n fre - quent external war , even though matrilineal descent corre- lates significantly with "North Ameri ca ." Furthermore , why i s des cent the independent variable and frequency of warfare t he dependent variable in the directi ons of strongest predic- t i on a ccording to lambda? Table 41 gives the lambda values where they occur in both directions between variables. By inspection , it appears t~at directional differences are less than , say , tho s e between variables i n the I nsular Pacific . Several of them are within fract ions of being perfectly mutually predictable , e . g ., frequent external war-attacki ng with official decision- maki ng to go to war . Such mutuality i s , I think , evidence that patterns of circular causality can be very tightly interwoven and may be i n the long run like a Gordian knot ~ From Table 40 one can dis cern fi ner geographical patterning within the continent . I n cell~' the Pawnee, Huron , Creek , Nat chez , and Aztec share a broad ecological and historical background of Eastern (or Southeastern) Woodlands , i ncipient or intensive agriculture , and paths of cultural diffusion . In cell Q there are 4 societies from the southern Plains and the Southwest, although their subsistence bases are differ- ent . I n cell Q, 5 of the 14 cases are either Northern Athapaskan or Algonkian footi-collectors , 3 are from Cali- fornia , 1 the Great Basin , and 2 (the Havasupai and Hui chol) are i n refuge areas . TABLE 41 . SIGNIFICANT CORRELATIONS WITH LAMBDAS I N TWO DIRECTIONS--NORTH AMERICA Pairs of Vari abl es . cp p "' r "' C Row attribute 1-4 Levels Beyond the Lo cal Community (1 4 ) wi th : High prest i ge for warriors ( 33) . 513 • 0 1 . 300 . 364 Elaborate rewards (35) . 476 . 01 . 100 . 083 High value for war ( 37 ) . 465 . 01 • 1 1 1 .. 385 Row attribute Patrilineal Descent (16 ) with : Frequent external war- attacking ( 22) -. 8 10 .01 . 66 7 . 750 Frequent external war--being attacked (23 ) -~ 756 . 0 5 . 500 . 667 Row attri bute Hereditary Headman (20) with : Frequent external war- attacking (22 ) . 378 . 05 . 286 , 375 Official decision ( 25) • 4 36 . 05 . 385 . 385 Row attribute Frequent External War- Attacki ng (22 ) wi th : Frequent ex~ernal war--being attacked t23) . 804 . 001 . 786 .727 Official decision ( 25) . 496 . 01 . 417 . 462 Military expectations I (29) . 439 . 02 . 385 . 273 High prestige (33) . 46 1 • 01 . 438 • 182 Elaborate rewards (35) . 665 . 001 . 643 . 61 5 High value for war (37) . 4 71 . 0 1 . 467 . 385 418 TABLE 41 (Continued ) Pairs of Variables cp p Ar A C Row attribute Frequent External War--Being Attacked ( 23) with : Military expectations I ( 29) . 665 . 001 . 500 . 600 High prestige (33) . 554 . 01 . 273 . 1 1 1 Elaborate rewards ( 35 ) . 686 . 001 . 556 . 636 High value f or war ( 37) . 586 . 01 . 400 . 538 Trophi es and honors (44 ) . 428 .05 . 125 . 364 Row attribute Military Expectations I (29) with : Hi gh value for war ( 37 ) . 432 . 05 . 200 . 385 Trophies and honors (44) . 786 . 001 . 727 . 750 Row att ribute High Prestige (33) with : Elaborate rewards (3 5 ) . 665 . 001 . 4 55 . 538 Expectations of v i olence ( 36 ) . 573 . 01 . 400 . 500 Hi gh value for war ( 37) . 532 . 01 . 300 . 462 Row attri bute Expectations of Violence (36 ) with : Hi gh value for war ( 37 ) . 529 . 01 . 500 . 455 6 . South Ameri ca . Not a ll of the 32 socie ties i n this reg i on are untouched by European contact . The Haitians and the Carib are t he re sult of the s l ave trade between Afri ca and the New World . The 32 are : 419 155 Yucatec Maya 166 Mundurucu 177 Tupi namba 156 Mi skit o 167 Cubeo 178 Botocud o 157 Bri bri 168 Cayapa 179 Shavante 158 Cuna 169 Jj_varo 180 Aweikoma 159 Goaj i ro 170 Amhuaca 181 Cayua 160 Haitians 1 71 I nca 182 Lengua 161 Callinago 172 Aymara 183 Abipon 162 Wa r rau 173 Siriano 184 Mapuche 163 Yanomamo 174 Nambicuara 185 Tehuelche 164 Carib 175 Trwnai 186 Yahgan 165 Saramacca 176 Ti mbi ra The soci eties are located by their ident ity numbers on map 6 . Despite the relatively high percentage of shared dis- tribution of phis (Table 31) between North a nd South America , t he s i gnificant patterns among the variabl es are notably dif- ferent in associa tions , dire ction , and strength of predicta- bili ty . Warfare in South America appears to be primarily o~fensive , and the attribute external war - attacking predicts a t t i tudes , expe ctat ions , and va l ues with regard to war and pre dicts tha t 18 per cent of the time the average community size i nvolved will be 100-400 persons (11) . The warfare variables pat t ern large ly i n i solation from political organi- zation (13) . In Figure 13, where there are s i gnifica nt asso- ciat ions with subjugation (40) and t r i bute (41), and both grouped together (30), the centralized polities i nvolved are the Saramacca and the I nca . Furthermore , polity i s not sig- nificant l y related to anything other than milit ary expe cta - tions and complex settlement pattern ; sett l ement pa ttern by i tself generate s association wit h military organization (age - sets , military socie t ie s , standi ng a r nie s ( 24 ); f ormal com- m nd (32) . Even when political organi zation i s expanded to 420 Map 6: South America 1~7 •163 oo 166 • .176 9170 175 • • 74 173 179 • Miles O 400 800 200 600 (adapted from Murdock and White 1969: 347) This map shows ";/J of the 32 societies classed as South American. 421 'Fig. 13 South America: direction and strength of lambdas, focused on 2-4 levels of polity (13). 422 i nclude 1 to 4 levels of j uri sdic t ioD ( 14 ) , the only warfare variabl e other than military expectations that it is signi f i - ca nt l y as s ociated with is high prestige for warriors (33) (Figure 14 ). I n a l arge sense , the warfare complex in the lower hal f of Figure 13 exi sts in isolation , att ribute s of war pre dictable primarily through other at tributes of war . For i ns t ance, military succe s s (38) i s a better predictor of the presence of authoritative military command ( 32) than is se tt lement pat tern (8 ) . It looks to me as though the warfare complex i n South America exi s t s independently, a nd i t is o~fensive warfare . Even t he relationship between offensive a~d defensive exte rnal war (2 2 and 23) , so strong in associ- at ion and predict ability i n North America , barely makes the . 05 level of significance i n South America . Tabl e 42 is the i ntercorrelat ion of frequent offensive war with military suc- cess ; i t shows who i s winning i n South America . In the 6 reg ions , t he a ssoci at ion between offens ive war and success i s s i gni ficant in 3 , predictable i n 2 , a nd strongest i n South America . The compa r i son looks like this : Frequent External War- Atta cking wi th Military Success (Expans ion ) cp p ;\ T C "r Ci rcum- Mediterranean . 433 . 05 0 0 . 188 Ea s t Eurasia . 4 ·10 . 0 2 0 . 250 . 168 South America . 538 . 01 . 1 00 . 1 82 . 2 89 423 @ .s,, .:-1 G @ @ @) Fig. 14 South America: direction and strength of lambdas, focused on 1-4 levels of polity (14). TABLE 42. SOUTH AMERICA : INTERCORREIJ1.TION OF MILITARY SUCCESS 424 AND EXTERNAL WAR-A'I'I'ACKING Territorial Expansion Present Territorial Expansion Absent Miskito Yucatec Maya Bribri Yanomamo Callinago Aymara Saramacca Nambicuara Mundurucu Timbira Jivaro Botocudo Inca Shavante Tupinamba Aweikoma Abipon Lengua Mapuche 10 9 Cuna Goa j iro Haitians Warrau Carib Cubeo Cayapa Siriano Trumai Cayua Yahgan 0 11 Q = 1.00 4 - .538 x2 = 8.684 p < .01 ).. = .. 100 )... = .1 82 C r 42 5 Earlier , I i dentifi ed South America as being part of a world- wi de androcentri c configurat ion , se condary t o Afri ca . Tha t androcentrism i s mani fes t i n the intercorrelations wi thi n the region concerned with i nternal war (21 ). Where there is frequen t internal war, 44 per cent of the time one can predict the presence of pat r ilocality ( 3 ), and 50 per cent of t he time one can predict t he presence of the expe cta- tion that violence is a s olut io n to problems (36 ). David Maybury- Lewi s ' hypothe s is that men ' s houses and conflict go t ogether may be correct , but i t appears to be l imited to South fune rica . Tha t is , t here i s no functional link between male groupi ngs and t he perpetuat ion and intensificat ion of attitude s and value s held about violence . Instead , histori- ca lly it has occurred i n one geographic region ; it i s truly a c1-J_l tural relationship , i . e . , arbitrary . Table 43 gi ves those attributes that are predictable i n both di rect ions . When compared to Table 41 , which pre- sents North American data, one can readily see tha t even in contiguous geographica l regi ons that are supposed to have such strong so ciocultural similari ty , the r e l at ionships among the warfare variables within each region are sharply dis- tinctive and have l ittle i n common. 42 6 TABLE 4 3 . SIGNIFICANT CORR.ELATIONS WITH LAMBDAS I N TTO DIRECT IO_S--SOUTH AMERICA Pai rs of Vari ab l es cp p Ar A C Row attri bute Polygyny ( 2 ) with : High val ue for war (37 ) . 537 . 02 . 375 • 444 Row attri bute Patrilocali ty ( 3) with: Frequent i nte rna l war ( 21) . 492 . 05 .444 . 375 Row attribute Complex Sett l ements (8 ) wi th : 1-4 levels beyond the l ocal c ommunity (1 4) .710 . 001 . 600 . 333 Military expectations I ( 29 ) . 518 . 01 . 364 . 41 7 Military expe ctat i ons II ( 30 ) . 567 . 01 . 455 . 14 3 Authori tat ive command ( 32 ) . 462 . 02 . 222 . 417 R01·1 attribut e Mean Com.muni t y Size 100- 400 Pers ons (11 ) with : Frequent i nternal war ( 21) -. 408 . 05 . 222 . 300 Row at tribute 2- 4 Levels Beyond t he Local Community (1 3) wi th : Tribute (41 .) . 802 . 001 . 500 . 667 Row attribute 1-4 Leve l s Beyond t he Local Community ( 14) with : Military expe ctat i ons I I ( 30) . 709 .00 1 . 500 . 571 Row attribute Frequent I nte rnal War ( 21 ) wi th : Hi gh value for war ( 37 ) . 481 . 05 . 250 . 333 Row attri but e Fre quent External War- Attacki ng (22 ) with : Rewar ds for warriors ( 35 ) . 756 . 01 . 500 • 667 Expe cta tions of vi olence (36 ) . 601 . 01 . 429 . 556 Military success (expansion)(38 ) . 5 38 . 01 . 182 • 1 00 Row attribute Age - Sets ... Armi es ( 24) .wi th : Authori tative command ( 32 ) . 514 . 02 .1 43 . 455 427 TABLE 43 (Co~t i nued) Pai rs of Vari ables p Ar A C Row att:cibute Military Expectations I ( 29) 1vith: Authoritative command (32 ) . 769 . 001 . 727 . 750 Hi gh prest i ge (33 ) . 664 . 01 . 600 . 636 Rewar ds for warri or ( 35 ) . 553 . 02 . 500 -333 Hi gh val ue f or war ( 37) . 424 . 05 . 143 . 333 Military success ( 38 ) . 722 . 001 . 667 . 600 Row at t r i but e Authoritat i ve Command (32) wi t h : Rewards for warri ors ( 35 ) . 645 . 01 • 571 . 500 Mili tary success (expansion ) (38 ) . 523 . 01 . 500 . 400 Row attribute High Prestige (33) with : Expe ctations of vi ol encP. (36 ) . 630 . 0 2 . 625 • 571 Mili tary success (e xpansi on) ( 38 ) . 574 . 01 . 500 . 444 Trophies and honors ( 44) . 516 . 02 . 455 . 250 Thi s concludes my presentati on of t he di s coveri e s of the ODYSSEY . I n t he f uture , I hope to e nl arge t he r egional sa□ples and run the i nt ercorre l at i ons agai n , to s ee if the preli minary f i ndings hold . The one weakness of the Standard Sampl e that I can see i s that it is not l arge enough t o make one conf i dent when do i ng regional analyses be cause of t he size of the Ns . I have commented earl ier t hat s i n ce t he regional subsampl es are a l ready similar , i t i s not the same thing as working with 30 societies spread out over t he wor l d . Nevertheless , I would l ike to enlarge the number f rom at least 1 or 2 regi ons , us i ng Murdo ck ' s World Sampl ing 428 Provi nces (1968 ), t o see if any discrepancies show up . I f t hey d o not , I will be more co~fident about my regio nal analyses . 42 9 FINAL REMARKS I t has been a long voyage. I have logged and charted my exploration of the world of war and found its nature to be intricate ly variable , shifting in time, space, and per- spective . I have argued against universa l explanations of warfare and for the hypothesis of variability due to factors of geography and history . I have performed quantitat i ve t ests that I think show that a pattern of relationships among var i able s that is a summary of da ta from cultures t hrougnout the world exists in the first place be cause of i nfl uences from particular geographical regions and yet does not exist 1"ithin any 1 regi on in the same form or strength . Further- ~ore , it is not permissibl e to transform variability into u:'.'liformity through the magic of evolutionary seriation . The diachronic qualities built into the Standard Sample make such pract ice questionable methodologically , without having to go i nto phi losophical di sputation . As Driver and Schuessler have done , I conclude that the ge ographical patterning and var iability in configurations support geographica l - historical explanations of warfare rather than psycho- f unct ional ones . However warfare in a given region gets started , it becomes part of a lattice of variables--which one s we may not know a priori-- that is probably in most ca ses self-perpe tuating . Unilinea l, one-directional , determinist ic 4 30 causal c _ains are simply too crude and clumsy to a c count for the data . Certainly polygyny i s part of a warfare syndrome- -in African external 1rmr and South American internal war . Cer- tainly food- production is related to warfare- -in Africa and t he I nsular Pacific. Of course political centralization is i mportant--but i t doe s not have any direct relationship with warfare a nywhere i n the world , and its re lationship to the sociocultural institutions of war is highly variable . The presence of some supra- communi ty organi zation--sodality , chiefdom , or state--does have a direct relationship with warfare - - but not j n East Eurasi a , the I nsular Pacific , or SoutD America . Moreove r , in t he remaining regions , the pre - dict i ve direction and strength differ . Yet indeed , "f rater- na l i ntere st groups " are i mportant, but on t he world scene the principle of f ormat ion is unilinea l de s cent and t he war - fare is i nternal . In Africa i t is patrilineal descent and t he warfare i s offensi ve external . In the Ci r cum- Mediterranean the principle may be unilineal des ce nt but it may also be greed, indivi dual prestige , and vengeance , and the form of war i s interna l . I n t he I nsular Pacif ic region , t he war i s also i nterna l , and the principl e may be greed- -or endogamous community organization . I n South Ameri ca , the war i s interna l and t he principle i s patrilocal residence . And I may say to myse l f , certainly a tt itudes toward violence an; primary--i n t 11e Circum- Med i te r ranean and South 431 Ameri ca . In the world , Africa , and East Eurasia t hey are more often dependent than i ndependent attr i butes . And yes , offens i ve warfare i s an i ndependent variable-- i n Afri ca , the I nsul ar Pacif ic, and South America . Of cours e the presence of hi gh i ndi vi dual prestige i s related to the f requency of warfare , but i n the worl d , Afri ca , and South Ame rica i t i s a dependent attri bute , while i n East Eurasi a and North A.merica i t i s an i ndependent attribute , and i n the Circum- Medite rranean i t i s significantly absent . Thus , some of my own universal generalizat~ons that I have offered throughout thi s dissertation a re sub j ect t o ge o- graphic patterni ng . Wha t more c ould I ask for? I f my basic argument is a cceptable , t hat the phenomen on of warfare can be understood only within the context i n which it occurs , and t hat the most i mportant component of the context i s human actors themselves , warfare studi es may proceed i n a theoreti- cal direction complete ly oppos ite from t hat of re cent re search . I have tried to show throughout t hi s work that the current myths in anthropology--about the nat ure of expla- nation i n social s cience, the nature of war , and the nature of the human beings who engage i n war--are just that , myths . And surely one point that we can agree upon i s that myths are not s cien ce , even though both try to make the world i n t elligible . 432 NOTES 1one is remi nded of a Herblock cartoon i n whi ch cit i- zens are standing i n l ine outside a j a il wi th t he s i gn "Police State " over i ts door , and one person remarks , "I t l ooks so ni ce and safe i n there ." 2As support fo r t hi s argument , note the followi ng com- ment by Le a ch (1 966 : 72 ): The development of spee ch i n Homo sapiens has compl e tely a l tered our nature . A goose can communi cate wi th ano t her goose by means of "ritualized " gestures , but t he ki nds of message it can t ransmi t are very narrowl y delimi ted : "Host i l j_t y " and 11 f r iendship " are onl y the obs e rver ' s l abels for s i mpl e tri ggered r esponses . I n cont r ast , human beings can say an i nf i ni t e number of t h i ngs i n an i nfi ni te number of di fferent ways ; responses are i ntrin- si ca lly unpr edictabl e ; polit icians and hi st orians are what they a r e precisely because n o man can eve r know what hi s "opponent " i s going to do next . 3one of t he probl ems with di sjunct i ons i s t hat we think of them i n either/ or te rms and t hat somehow one side uust be completely free of the other , e . g ., ob j ectivit y- s1,;_bj e ct i vi ty , f a ct - val ue , de t ermi ni sm-inde t ermini sm , rni nd- body , descri ption- explanat i on . What must be understood is that , while di s j unc t ions are probably universal in human language , the t wo halves of the dis j unct ion cannot be taken apart and still have any meaning . I t i s a l ogica l i mpossi - bi l i ty . 4Thi s business about l ife process could be ca l l ed t he Natu:cal Law of Growth and Development , which i mplies poli ti- cal as uell as physical attributes . Nevertheless , i f taken as true , it is conceivable that the a ccommodat i on of the ea1·th is coming to an end in terms of nrnnbers , and q_uali ta- ti ve changes do not appear to be happening . 5Popper mentions Ernpedocles ' theory of evo l ut i on , Parrnenides ' theory of an unchanging uni verse , and astrologi - cal theories of planetary influences ( cf . Newton ' s the ory of gravity and the l unar theory of the tides) . 433 6The use of "risky " may be misleading . Hanson (1958) no es tte difficulty of applyi ng falsification tests to l aws of classical physics . It i s aln,ost impossi ble to i magine situations in which the laws woul d not appl y , e .g . , objects s l owing down as they fall , or l ead balls falling faster than feathers in a vacuum . Note how fasc i nated peopl e are with weightless astronauts in space . I find f loating liquids especial ly odd . While some philosophers are impressed by the i nability to fals i fy and would deny statements of dynamics their law status , reducing them to convent i ons , rules , or def i nitions , Hanson demurs , arguing that our i nability t o i magine situat ions in which the dynamic laws do not apply is due not to t he logical status of the l aw but to the vast i nventory of confirmat i ons and our patterns of thinking i n dynami cal terms (1 958 : 93 ) . 7 11 ... every two materi al objects attra ct each ot her with a f orce proportional to the product of t heir masses , a nd i nvers ely proport i ona l to t he square of the di_stance between them " (Gamow 1962 : 44) . 8Because i ts i ncreased efficiency produces a surplus free i ng some food-producers from subsi stence a ct ivity . "The ratio between the amount of energy diverted to foo d produc- tion and the amount of energy availabl e for other a ct i vities i s one of the most important cultural variables " (Harri s 1 97 1 : 203) . 9sic . I nnovat ion seems more effectively l imi ted to change by a cons cious agent . Re combi nation and mutat ion i n biol ogy i s , of course , not conscious . But i n t he re a l m of culture i ndividuals do make conscious de cisions and choices a oout change all the time. The di st inction is i mportant . 10Note Chagnon ' s descriptions of 2 Yanoma~o brothers , one skept ical and caut ious , the other subscribing to the i deal of a super- warrior (1 968b) . 11 chagnon (1968b : 74- 75 ) describes a case of child neglect : a 2- year- ol d nearly starved because the mother t s milk had dried up and she , well-fed and pl ump , refused t o give it other foods because "it did not know how to eat other foo ds ." The sex of the child i s no t clearl y specified : Chagnon refers to t he child as "it " and "he ." 12Dival e notes that , f or the U. S. A. , the s ex ratio of live births is 104 males to 100 females ; f or We stern Europe i t i s 105 : 100 ; for Japan i t is 109 : 100 (1 970 :1 1) . 434 1 3•J1he prominent American figure i n resea rch on phys io- logical e nvironment and concept ion has been Dr . L. B. Shettles . He has found , for i nstance , hat out of patients acting ou his findings and suggestions for practical appli- cation , 23 of 26 couples who wanted to conceive boys were successful , and 19 of 22 couples who wanted girls were suc- cessful (Shettles 1970 ; Rorvik and Shett les) . 14rn 1970, Divale remarks tnat he is putt ing his war- fare syndrome together in the master ' s t hes i s , which includes case studies of the Yanomamo, t he Sioux, the Ibo, the Kapauku , the Central Eskimo , and t he Yir Yoront. 15While i t would not be polite to quote it here , as evidenc e that the Indian warrior ' s expectations of rewa rd persisted at least through World War II, I recommend to the reader Patton ' s address to his troops , in "My Favor ite Gen- eral" in Dwight Macdonald ' s Memo i rs of a Revolutionist, Meridian Books, 1957. 161 have based the position of population pressure i n the chain upon Harris ' assumption that in the long run " the rat e of population growth has always tended to increase in response t o new opportuni ties for such growth '' (1971 : 223). 17[Culture] is a plan for behavior , not behavior i tself ; it is arbitrary , learned , and shared . In addi - tion , cultur e i s adaptive . Human beings co pe with their natural and social enviror1ment by means of their tradi- tional categorie s and plans . Unlike other ani mals , if we are cold we can invent a coat and pass along the te ch- nique t o our descendant s . Culture allows for rapid adap- tation because i t is fle xible and permits the invention of new strategies- - a lthough change ofteu appears to be painfully slow to those who are in a hurry for it . By the same token , the adaptive nature of culture accounts for the enormous variety of the world ' s distinct societies (Spradley and Mccurdy 1971 : 4) . 18Jamil Nammour (1973) has worke d on the difficulties of those who want " to pry language off the world . " I have borrowed and adapted this singularly effective phrase . 19Inflation is a fine and familiar example of circular causation , yet American economists in charge of moderating it appear to be usi ng only the i nadequate tools of equilibrium theory. One of the beauties of Myrdal ' s theory is that it does operate in the real world . 435 20The following summary of quant ifica tion schools i n anthropology is by Driver (1 965 : 324- 45 ; Chaney 1973 : 1359-60) . The California and Yale Schools The California school , represented by A. 1 . Kroeber , h is pupils and associates , used proximity coefficient s first to determine geographical clusterings of ethnic units ( tribes , tribelet s , villages , bands , societies , and any other culture- bearing group of people ) and se cond to reconstruct the history of these ethnic units , of their total cultures, or of a r estricted part of their cul- tures . I n every case t he uni verse of i nvestigation was limited to an area of le ss than cont i nent al s cope . In most of these studie s the ethnic units were treated as variables and they were compared i n terms of t he amount of culture trait i nventory shared . The conclusions arri ved at were regarded as applyi ng only to a restri cted area of each s tudy , and no attempt was made to establish universal generalizations about the culture of the entire human species . The Ca lifornia school stemmed from Boas . The Yale school, represented by G. P . Murdo ck , J . W. M. Whi ting , and their pupils . and assoc i a tes , aimed at estab- lishing worl d- wi de generalizations about all ethnic uni ts from sampl es thought to be representative of the whole . These studies treat subject units (culture traits , ele- ments , compl exes , components ~ themes , pat terns , and any other parts of total culture ; as variables and compare them i n terms of the number of e thni c units sharing the various combinat ions of sub j ect units . The Yale school stemmed from Tylor , Stunner , and Keller . The California method was largely empirical in that it was not bolstered by formal postulates , theorems, or propositions . However , i t i mplicitly relied on t he gen- eral principle that cont i nuity of geographical distribu- t ion resul ted from geographical factors , di ffus i on , migration , and other geographico- hi storical pro cesses . It sometimes employe d the age- area hypothesis , which assumes a perfect correlation between size of geographi - cal area and age of the subject unit distributed over the area . The California method has been frequent ly labeled historical , but i t was inferential , undocumented hi story . The Yale school , on the other hand , was l argely postula- tional , in that considerable functional or psychologica l theory was f orm1J.lated in advance , and only the data rel- evant to the theory collected . Murdo ck ' s Social Struc- ture ( 1949 ) i s the best known exarnp~Le of this method and 436 Whiti ng ' s psychological studies (e . g ., Whit i ng and Child , 195 j) foll ow similar procedures . However, when Driver blunt l y asked Murdo ck a f ew years ago if every theorem a nd propos ition i n hi s Social Structure was formulated i n a dvance of t he colle ct i ng of data and printed wit hout a l te rat io r.:. after the data "\vas col lected , Mur c.. ock gave a negative r epl y ; the postulates were modified i n light of what the empirica l data revealed . Therefore the dichotomy of postulat ional versus e mpi r ical (po s itive) method , so c harmi ng to s ome philosophers , doe s not fit t he reality of actual research methods , in ethnology , and probably n ot in any other behavioral science . The Yale method has been labele d evol ut i onary because i t i nterprets i t s correlations as causal sequences repli- cat ed over a nd over again among societies thought not to be connected histor i ca lly . It ha s used te s ts of s i gnif- icance , especia lly chi- square , much more ofte~ than the Ca lifornia Schoo l . 21 Raoul Na.roll , "Warfare , peaceful i ntercourse , and t errit oria l chance : A cross - cultura l survey ," mimeographed , 1964 . Na.roll 1966 is a highly condensed version . 22It is also apparent l y being publi shed by Naroll , Vern R. Bu llough , and Frada Na.roll as Mi litary De terren ce i n Hi story : A Pilot Cross-Hi storical Survey , State Universi ty of NeK York, 1974 . I n Books in Print 1973 there is a second entry for these authors , Military De t errence i n History : A Statisti cal Survey , State University of New York , 1973 . I have seen neither of these publications . 437 APPENDICES 438 APPENDIX A STANDARD CROSS~CULTURAL SAJ1PLE Each society in Murdock and White ' s Standard Cross- Cultural Sample (1969) is listed by i dent ity number and name , followed by the target group, it s location , and the date to which the ethnographic descript ions apply, exactly as given by Murdock and White . 1. Nama Hottentot . The Gei/Khauan t ribe ( 27° 30 ' 3, 17°E) reconstructed for 1860 , just prior to their decimation and loss of independence in the Herero War . 2 . Kung; Bushmen . The Agau Kung of the Nyae Nyae regj_on (19 5o rs , 20°35 1 E) i n 1950 . 3 . Thonga . The Ronga subtribe around Lourenco Marques (25°50 1 S , 32°20 ' E) i n 1985 . 4 . Lozi or Barotse . The ruling Luyana (1 4° -1 8°20 ' 3 , 22° - 25°E) in 1900 , at the height of Barotse political expansion . 5 . Mbundu or Ovimbundu . Bailundo subtribe (12°15 1 S , 16°30'~ in 1890 , just prior to Portuguese conquest and mis s ion- ization . 6 . Kongo . 17°S , ·15°E . 1900 . [Substitution for the Suku . ] 7 . Ila . 16°S , 27°E . 1957 . [Substitut i on for the Bemba .J 8 . Nyakyusa . The Nyakyusa around the towns of Mwaya ( 9°35 rs , 34° 10 ' E) and Masoko (9°20 1S , 34°E) i n 1934 . 9 . Hadza of Kindiga . The small Hadza tribe as a whole (3° 20 '-4° 10 1 S , 34°40 1 - 35°25 1E) i ll 1930, when still unacculturated . 10 . Luguru or Waluguru . The Luguru of west central Morogoro District (6°25 ' -7°25 1 S , 37°20 1 - 38°E) , in 1925 , the last date of the traditional political org nization . 439 11 . The kikuyu of the Metume or- Fort Hall district (0° 4o rs , 37°10'E) i n 1920 , prio r to intensive a cculturation~ 12 . Ganda or Ba&anda . The ganda of Kyaddondo district (0°20 ' N, 32 30 ' E) in 1875 , just prior to the founding of Kampala and the initiation of signifi cant admini strat i ve changes . 13 . Mbuti Pygmies . The E;~ml u net - hunters of the I turi Forest (1°30 ' -2°N , 28 15 '-28°25 ' E) i n 1950 . 14 . Nkundo -Mango . The Mango of the Ilanga subtri be (0°15 '-1° 15 1 S , 18°35 '-1 9°45 ' E) i n 1930 . 15 . Banen or Banyin . The Ndi ki subtribe (4°35 '-4°45 1N, 10°35 '- 11°E ) i n 1935 . 16 . Tiv . The Tiv of Benue Province ( 6~30 '-8°N , 8°-1 0°E ) i n 1920 , pri or to extensive organizational changes wrought by the British . 17 . Ibo or Igbo . The Eastern and Peripheral subgroups of the Isu-Ama division of the Southern or Owerri I bo ( 5°20 1 - 5°40 1 N: 7° 10 '-7°30 ' E) in 1935 . 18 . Fon or Dahomeans . The Fon i n the vi cinity of Abomey (7°12 1 N, 1°56 ' E) i n 1890 , prior to the conquest of the Dahomean kingdom by the French. 19 . Ashanti . The Ashant i of the state of Kumas i ( 6° --8° N, 0° - 3°W) i n 1895 , just pri or to British conquest . 20 . Mende . The central Mende around the town of Bo ( 7° 50 1 N, 1 2° W) in 1 94 5 . 21 . Wolof or Quolof . The wo l of of Upper and Lower Salum i n the Gambi a ( centering on 13°45N , 15°20 1 W) i n 1950 . 22 . Bambara . The Bambara along the Niger River from Segou of Bamako ( 12° 30 1 - 13°N , 6° - 8°W) in 1902 . 23 . Tallensi . The small Tallensi tribe as a whole (10°30 '- 10045 ' N, 0°30 '-0°50 ' W) i n 1934 . 24 . Son&hai . The s onghai of the Bamba or cent ral division ( 16 -1 7° 15 1 N, 0° 10 1 E- 3° 10 1 W) in 1940 . 25 . Fulani . The Alijam and Dege.ciji subgroups of Wodaabe Fulani around Adan and Damergou i n Niger (1 3° -1 7°N , 5° -1 0°E) in 1951 . 440 26 . Hausa . The Zazzagawa Hause ( 9°30 '-11 °30 ' N, 6° - 9~ E) in 1900 , just prior to the aaver:t of British r ule . 27 . Kanuri of BQr nu . 12° N, 13°E . 1870 . [Substitution for Mas sa . ] 28 . Azande . The Azande of the Yambio chi efdom (4°2 o r_ 5°50 ' N, 27° 40 '-28°50'E ) i n 1905 , j us t prior to Bri t ish conques t a nd the collapse of the Avongara politica l system. 29 . :F'ur . The Fur of western Darfur around J ebe l Marra ( 13°3o rN, 25°30 ' E) in 1880 , prior to effective Egypt i an subjugat i on . 30 . 0toro Nuba . The 0toro of the Nuba Hill s (1 1°20 ' N, 30°40'E) i n 1930 , prior to substantial migra tion i nto the plains . 31 . Shill uk . The politicall y unified Shilluk as a whole (9° -1 0u30 ' N, 31° - 32°E) in 1910 . 32 . I ngas sana . 33 . Kafa . The poli t ically unified Kafa as a whole ( 6°50 '- 7045 1N, 35°3o r- 37° E) i n 1905 . [The ethnographic materi- als for this soci ety are i n German . The re are no sub- stitutes offered i n the Standard Sampl e . Therefore , there is no codi ng for warfare dat a . J 34 . Masai . The Kisonko or Southern Masai of Tanzania ( 1°30 r- 5°30 ' S, 35c - 37°30 'E) in 1900 . 35 . Kons o . The Konso of the town of Bus a (5°15' N, 37° 30 ' E) 36 . Somali . The Dolbahanta subtri be (7°-11° N, 45°30 '-49°E) i n 1900 , subsequent to the earlie st de s criptions but prior to t he l ater and fuller ac counts . 37 . .A_mhara . The Arnhara of the Gondar di strict ( 11 ° - 14 ° N, 36~- 38°30 ' E) i n 1953 . 38 . Bi s ha rin . 20°N , 35° E . 1930 . (Sub s t i tution for t he Bog a . ] 39 . Nubians . The Kenuzi or northernmost branch of the Bara bra or Nile Nubi ans (2 2°-24° N, 32° - 33°E) i n 1900 , j us t prior to the i r di splacement by the firs t Aswan dam. 40. Teda . The Teda of Tibest i (1 9° - 22°N , 16° -1 9°E) i n 1959 . 441 41 . Tuareg . ~Che Ahaggaren or '.ruareg of Ahaggar ( 21° - 25° N, 4° - 9°E) i n 1900 , pri or to the Fre nch military occupat i on of the Sahara . 42 . Riffians . The Riffians as a who l e (34° 20 '-35°30 1N, 2°30 '-4° W) in 1926 . 43 . Ancient Egypt i ans . 30°N , 31°E . 1200 B. C. [Substitu- t i on for 20th century Egypt ians from the town of Silwa . ] 44 . Hebrews . The kingdom of Jud.ah ( 30°30 1 - 31°55 ' N, 34° 20 1 - 35° 30 ' E) i n 621 B. C., the date of promulgation of the Deuteronomic l aws . 45 . Babylonians . The city and environs of Babylon ( 32°35'N , 44°45 ' E) i n 1750 B. C., at the end of the reign of Hammurabi . 46 . Rwal a . The Rwala Bedouin of south central Syria and northeastern Jordan ( 31° - 35°30 1N, 36°-41°E ) i n 1913 . 47 . Turks . The Turks of t he northern Anatolian plateau ( 36°40 ' - 40°N , 32° 40 1 - 35°50 ' E) i n 1950 . 48 . Ghe& . The Mountain Gheg of northern Albani a (4 1°20 1 - - 42 N, 19° 30 1 - 20°31 ' E) i n 1910 , j ust prior to the expul- s i on of t he Turks in the two Bal kan wars . 49 . Romans . The Romans of the city and environs of Rome (41°50 1 N, 13°30'E) i n A.D . 110 , t he t welfth year of Traj an ' s reign at the approxi mate zenith of the i mperi a l peri od . 50 . Basgues . The mountai n village of Vera de Bidasoa (43°1 8 ' N, 1° 40 1 W) in 1934 . 51 . I rish . The Irish of County Clare ( 52° 40 '-53°10' N, 8° 20 1 -1 0°W) i n 1932 . 52 . Lapps . The Konkama Lapps of Karesuando ~arish i n northern Sweden ( 68° 20 ' - 69°5 ' N, 20° 5 '-23 E) in 195 0 . 53 . Yurak Samoyed . The Tundra Yurak (65°-71° N, 41° - 62°E ) i n 1894 . 54 . Rus s i ans . The Great Russi ans of t he peasant village of Viriat i no ( 52°40 ' N, 41° 20 ' E) i n 1955 . · 55 . Abkhaz . The smal l Abkhaz t r i be as a whole (42°50 1- 430 25 1N, 40° - 41°35 'E) i n 1880 . 442 56. Armenians . The Armenians in the vicinity of Erevan (40°N , 44°30 1 E) in 1843 . 57 . Kurd . The Kurd of the town and envi rons of Rowandux (36~30 1 N, 44°30 ' E) i n 1951 . 58 . Basseri . The nomadi c Basseri (27° - 31°N , 53° - 54°E) in 1 958 . 59 . Pun~abi . The western Punjabi of the village of Mohla ( 32 30 ' N, 74°E) in 1950. 60 . Go nd . The Hill Maria Gond ( 19°15'-20° N, 80°30 1 - 80°20 1 E) i n 1938 . 61. Toda . The small Toda tribe as a whole i n 1900 (11°-1 2°N , 76° -77°E) . 62 . Santal . The Santal of the Bankura and Birbhum districts of Bengal (2 3°-24°N , 86°50 '-87° 30 ' E) i n 1940 . 63 . Uttar Pradesh . The village of Senapur i n the small kingdom of Dobhi Ta1uka ( 25°55 1N, 83° E) i n 1945 , prior to a major shift in the tradi tional power base. 64 . Burusho . The Burusho of Hunza state ( 36°20 1 - 36°30 1 N, 74°30 1 - 74°49 rE) in 1934 . 65 . Kazak . The Kazak of the Great Horde ( 37° - 48°N , 68° - 81°E) i n 1885 . 66 . Khalka Mongol s . The Khalka of the Narobanchin temple territor y (47c - 47u20 1 N, 95°10 1 - 97°E ) i n 1920. 67 . Lolo . The i ndependent and relatively unac culturated Lolo of the Ta liang Shan mountai ns ( 26°-29°N , 103°- 104°E) i n 191 O. 68 . Lepcha. The Lepcha in the vicini ty of Lingthem i n Sikkim (27° - 28°N , 89°E ) in 1937 . 69 . Garo . The Garo of Rengsang&ri and ne ighboring i nter- marrying villa es ( 26° N, 91 E) i n 1955. 70 . Serna Naqa . 26°N , 95°E . 191 0 . [Substi tut ion for the Lakher . J 71 . Burmese . '.L'he village of Nondwin i n Upper Burma ( 22°N , 95°40 ' E) in 1965. 443 72 . Palaun~ . 23°N, 97°E . 1920 . [Subst i tut ion for t he Lamet . J 73 . Viet namese . The Tonkinese or North Viet namese of the delta of the Red River ( 20° - 21°N , 105°30 '- 107°E ) in 1930 . 74 . Rhade . The Rhade of the viil age of Ko - s i er on the Darlac pl ateau (1 3°N , 108°E ) i n 1962 . • 75 . Khiner or Camb odians . The city of Angkor (1 3°30 ' N, 103°50 ' E) , the capita l of the old Khmer ki ngdom at its height , i n 1292 . 76 . Si amese or Central Thai . The Central Tha i village of Bang Chan (1 4°N , 100°50 ' E) about 1955 . 77 . Semai or Senoi. 4°N , 102°E . [Substitution for the Semang ; dat a extremely thin , while tho se f or the Semai are better . Al s o , t he Semai are clearly a no n- warring, nonviolent group and therefore are necessary to augment the re l at ively small number of such peoples in the Standard Sample . ] 78 . Nicobarese . The Nicobare s e of the northern i slands of Car Nicobar , Chowra , Teressa , and Bompoka (8° 15 '-9° 15t N, 92° 40 1 - 93°E ) i n 1870. 79 . Andamanese . The' Aka- Bea tri be of South Andaman (1 1°45 '-1 2°N , 93° - 93°10 ' E) i n 1860, prior to s i gnificant a cculturation and depopul at i on . 80 . Ve dda . The Dani gal a group of Forest Ve dda (7° 30 '-7°N , 81°-81° 30 ' E) i n 1860 . 81 . Tanala . The Menabe subtri be (2 2°S , 48°E ) i n 1925. 82 . Negri Sembi J_an . The district of I nas ( 2°30 '-2°40 1 N, 102° 10 1 -1 02°20 1 E) i n 1958 . 83 . J avanese . The town and envi rons of Pare i n central Java (1°43 1 S , 11 2°13 ' E) i n 1954 . 84 . Balinese . The village of Tihingan i n the district of Klunghung (8°30 1 S , 105°20 1 E) i n 1958 . 85 . I pan or Sea Dayak . The I ban of the Ul u Ai group ( 2° N' 11 2° 30 r -1 1 3° 30 IE) i n 1 958 . 86 . Bad j au . The Bad jau of southwestern Tawi-Tawi and ad j a cent i slands of the Sulu Archipe lago ( 5;:;N , 1;:?0°E ) in 1963 . 444 87 . Toradja . The Bare ' e subgroup of east ern Toradja (2°S , 12 1° E) i n 1910 . 88 . Tobelorese or Tobelo . The Tobelore s e as a whole ( 2° N, 1 28° E) in 1 900 . 89 . Alorese or Abui . The village complex of Atime lang in north cent ral Alar (8°20 ' S , 124° 40 ' E) i n 1938 . 90 . Tiwi . The Tiwi of Bathurst and Melville I s l ands as a whole (11°45 ' S , 130° - 132°E ) i n 1929 . 91. Aranda or Arunta . The Arunta Mbainda of Ali ce Springs ( 23°30 ' -2 5°S , 132°30 '- 134°20 ' E) in 1896 . 92 . Orokai va . The aiga subtribe (8° 20 '-8°40 1S , 147°50 1 - 1480 10 1E) i n 1925 . 93 . Ki mam . The village of Bamo l i L northeast central Frederick Hendrik I s l and or Kolekom (7° 30 ' S , 138° 30 1E) i n 1960 . 94 . Ka oauku . The village of Botuke bo in the Kamu Val_ley (46 S, 36°E) i n 1955 . 95 . Kwoma . The Hongwam subtri be (4° 10 1 S , 142°40 1E) i n 1937. 96 . Manus . The vill age of Peri ( 2° 10 1 S , 147°E ) i n 1929 . 97 . New Ire land . The vil l age of Lesu ( 2°3o rs, 151°W) in 98 . Trobrianders . The i sland of Kiriwina (8°38 1 S , 151°4 1 E) 99 . Si uai or Motuna . The northeastern Si uai of southern Bougai nville (7°S, 155°20 ' E) in 1939 . 100 . Tikopia . The small i sland of Tikopia as a whole ( 1 2° 30 ' S , 1 68° 30 ' E) i n 1 930 . 101. Pentecos t . The village of Bunlap and ne i ghbo r ing i nt er- marrying pagan villages in sout heastern Pentecost Island (1 6°S , 168°E) in 1953 . 102 . Mbau Fijians . The i sland of Mbau off the eas t coast of Vi ti Levu (1 8°S , 178°3 5 ' E) in 1840 , the approxi mate date 103 . Ajie . The petty chiefdom of Nefe ( 21°20 1 S , 165°40 ' E) reconstructed fo r 1845, prior to strong European i nfluence . 445 104 . Maori. Tr1e Nga Puhi tribe of the northern i sthmus (35° 10 1 - 35°30 1 S , 174c -1 74°20 rE) i n 1820 , prior to Euro pean settlement and mi ssioni zat ion . 105 . Marquesans . The Te- i ' i chiefdom of southwestern Nuku Hiva I sland (8°55 ' S , 140°10 ' W) about 1800 , at ab out the time of the ea liest reli able d~scriptions . 106 . Samoans . The kin&dom of Aana i n western Upolu I sland ( 13°48 1 - 14°S , 171 54 1 -1 72°3 1W) i n 1829, pri or to the military defeat of Aana a nd the beginning of i n tens ive European contact . 107 . Gilbertese . The northe rn Gilberte s e of Makin and But i r itari islands ( 3°30 ' N, 172°20 1E), reconstructed f or about 1890 . 108 . Mar shalle se . The atoll of Jaluit (6~N, 165°30 ' E ) i n 1900 , the mean date of the early German ethnographers . 109 . Trukese . The island of Romonum or Ulalu (7°24 ' N, 151°40 1 E) in 1947 . 110 . Yapese . The i sland of Yap as a whol e ( 9°30 1 N, 138°1 0 1 E) i n 1910 . 11 1 . Palauans . The village of Ulimang in northern Babelthuap Island ( 7° 30 1 N, 134°3 5 ' E) i n 1947. 112 . Ifu&ao . The Central and Kiangan Ifugao ( 16°50 1 N, 121 1o rE) i n 1910 . 113 . Atayal. The Ataya l proper (excluding the Sedeq ) as a whole (23°50 1 - 24v50 1 N, 120° 20-1 20° 50 1E) about 1930 , when the abori ginal culture was s till relative l y i ntact . 114 . Chi nese . The villa&e of Kaihsienkung in northern C~ekiang ( 31°N , 120 5 ' E) i n 1936 . 115 . Manchu . The Aigun di strict of northern Manchuria ( 50° N, 125° 30 ' E) in 1 91 5 . 116 . Koreans . The village of Sondup ' o and town of Samku Li on Kanghwa Island ( 37°37 ' N, 126° 25 ' E) i n 1947 . 117 . J apanese . The village of Niiike i n Okayama prefecture ( 34°40 ' N, 133°48 1 E) i n 1950 . 11 8 . Ai nu . The Ainu of t he basi ns of the Tokapchi a nd Saru rivers in southeast ern Hokkaido (42°40 ' -43° 30 ' N, 142° - 1440E) re cons tructed for about 1880 . 446 11 9 . Gilvak . The Gilyak of Sakhalin I sland ( 53° 3o r- 54°30 ' N, 14 1~ 5o r-1 43° 10 1 E) i n 1890 . 120 . Yukaghir . The Yukaghir of the Upper Kolyma River ( E3°30 1 - 660N, 150° - 157° E) i n 1850 , prior to marked depo pul at ion . 121 . Chv_kchee . The Reindeer Ctukchee (63° - 70°N , 171 °W- 171°E) i n 1900 . 122 . I ngalik or Tinneh . The vill age of Sha geluk ( 62°30 1 N, 159°30 ' W) , reconstru cted for 1885 , just prior to mis- sioni zat i on . 123 . Aleut . The Unalaska branch of the Aleut (53° - 57°30 1 N, 158°-170°W) about 1800 , prior to int ensive ac cultur- at i on . 124 . Co p~er Eski mo. The Copper Eskimo of the Arctic ma i nland (6t 40 1- 69°20'N , 108° -117° W) i n 1915 . 125 . Montagnais . The Montagnais of the Lake St . John and ~i stassini bands (48°- 52° N, 73° - 75°W) in 1910 . 126 . Micmac . The Micmac of the mainland (43°30 1- 50°N, 60°-66°W) i n 1650. 127 . Saulteaux . The Northern Saul teaux of the Berens River band ( 52°N , 95°30 ' W) i n 1930 . 128 . Sl ave . The Slave in t he vicinity of Fort Simpson (62°N, 122°W) i n 1940, j ust prior to the heavy a cculturat ion followi ng World War II . 129 . Kaska or Eas tern Nahani . The Kaska of t he Upper Li ard River (60°N , 131°W) , reconstructe d for 1900 , jus t prior to i ntensi ve missioni zat ion . 130 . Eyak . The s mall Eyak t r i be as a who l e (60°-6 1° N, 144° - 1460W) in 1890 , pri or to f ull a cculturat ion . 131. Hai da . The village of Masse t ( 54° N, 132°30 1W), re con- s tructed for 1875 , i mmediately pr i or to mi ss ionization . 132 . Bellacoola . The central Bellacoola a long the l ower Bella Co ol a River ( 52° 20 1N, 126° -1 27°W) i n 1880 . 133 . Twana . The small Twana tribe a a whole (47°20 1 - 47030 ' N, 123° 10 1 -1 23° 20 1W), reconstructed for 1860 , prior to mi ss ionization . 134 . Yurok . The village of Tsurai (4 2° 30 1N, 124°W) i n 1850 . 447 135 . Pomo . The Eastern Pomo of Clear Lake ( 39° N, 123°W) i n 1850 , prior to the i nrush of European se t tlers . 136 . Yokuts . The Lake Yokuts (35°10 ' N, 11 9°20 ' W) , prior to t he influx of sett le ro following the gold rush . 1 37 . Paiut e . The Wadadi ka or Harney Valley band of Northern Pai ute (4 3° - 44° N, 11 8° -1 20° W) , reconstruct ed fo r about 1870 , just prior to the establishment of the re se r - vation . 138 . Kl amath . The Klamath tribe as a whol e (42° - 43°15 ' N, 121°2orw) i n 1860 . . 139 . Kutenai . The Lor.rnr Kutenia (48°40 '-49° 10' N, 11 6°40 1W) i n 1890 . 140 . Gros Ventre or Atsina . The homogen ous Gros Ventre as a whole (47° - 49°N , 106° -11 0°W) i n 1880 , shortly prior to mi ssioni zat i on and the di sappearance of the buf'falo . 14 1 . Hi datsa or· Minitari . The village of Hidatsa (47° N, 101° W) , reconstructe d f or 1836 , prior to depopulat ion i n a severe small pox epi demi c. 142 . Pawnee. The Ski di or Skiri Pawnee (42°N , 100°W) , reconstructed for 1867 . 143 . Offiaha . The Omaha tribe as a who l e (41°10 1 -41°40'N, 96°-97°W ) in 1860 , prior to the disappearance of the buffa l o . 144 . Huron or Wendot . The Attignawantan (Bear Pe ople ) and Attigneenongnahac (C ord Peopl e ) tribes of the Huron Confederacy (44° - 45°N , 78°-80°W) i n 1634 , the dat e of the beginni ng of Jesuit mi ss i onary a ctivi ty . 145 . Creek or Mu~kogee . The Upper Creek of Alabama ( 32°30 1 - 34020 rN, 35~30 1 - 86° 30 ' W) i n 1800 , pri or to Tecumseh ' s rebellion and removal to Oklahoma. 146 . Natchez . The polit i cally i ntegrated Natchez as a whole ( 31°30 1 N, 91° 25 1 W) i h 1718 , t he date of the arrival of the first missionaries and ethnographers . 147 . Comanche . The Comanche as a whol e (30°- 38° N, 98° -1 03°W) in 1870 , just prior to pacif icat i on and remova l to Okl ahoma . 448 148 . Chiricahua Apache . The central band or Chiri cahua proper ( 32°N , 109°30 ' W) i n 1870 , i mmediately prior to the reservation peri od . 149 . Zuni . The village of Zuni (35° - 35°30 1 N, 108°30 1 - 109°W) i n 1880 . 150 . Havasupai . The smal l Savasus>ai tribe as a who l e (35°20 1 - 36° 20 1 N, 11 1°20 '- 11 3 W) i n 1918 . 151. Papago . The Archi e Papago near Seels , Ari zona ( 32° N, 11 2°W) , i n 1910 . 152 . Hui cho l . The small Huichol tribe as a whole ( 22°N , 105° W) i n 1890 . 153 . Aztec. The city and environs of Tenochtitlan ( 19°N , 99° 10 'W) in 1520 . 154 . Popoluca . The Sierra Poruluca of t he town and vi ci nity of Soteapan ( 18°15 1N, 94 50 1W) i n 1940 . 155 . Yucate c Maya . 18°N , 90°W . 1520 . [ Subs t i tut i on f or the Quiche , for whom the warfare material i s severely deficient . ] 1 56 . Miskito . The Mi skito in the vicinity of Cape Gracias a Dios (15°N , 83° W) i n 192 1 . 157 . Bri bri . The Bribri tribe of the Ta l amanca na tion (9°N , 83°15 ' W) i n 1917 . 158 . Cuna or Tule . The Cuna of the San Bl as Archipelago ( 9° - 9° 30 ' N, 78° - 79° W) i n 1927 . 159 . GoaJ i ro . The homogeneous Goaj iro tribe as a whole ( I! 30 1 -12°20 1 N, 71° - 72°30'W) in 194 7 . 160 . Haitians . The Haitians of Mi rebal ai s (1 8°50 ' N, 72°10 ' W) in 1935 . 161. Callinago or I sland Carib . The Ca llinago of the i sland of Dominica (1 5°30 ' N, 60°30 ' W), re constructed fo r 16 50, shortly prior t o mi ssioni zat ion. 162 . Warrau . The Warrau of the Orinoco de lta (8° 30 '-9° 50 1 N, 60°40 ' - 62°30 1W) i n 1935 . 163 . Yanomamo . The Shamatari subtribe a.round the village of Bisaasi- t eri · ( 2°-2°45 ' N, 64°30 '-65° 30 ' W) i n 196 5 . 449 164 . Car ib . The Cari b along the Barama River i n British Gui ana (7°10 1- 7°40 ' N, 59°20 '-60°20 1W) i n 1932 . 165 . Saramacca . The Saramacca group of Bush _-egroes i n the upper basin of the Suriname River ( 3°-4°N , 55°30 1 - 56° W) in 1928 . 166 . JViundurucu . The sayan~a- dweJ-l ing; Mundurucu of the Rio de Tropas drai nage ( 6 - 7 S , 56 - 57 W) , re constructed for about 1850 , prior to the period of i ncreasing ass i mi - l ation . 167 . Cubeo . The Cubeo of the Caduiari River (1°-1°5o•N, 70° - 72°W ) i n 1939 . 168 . Caiapa . The Caya~a i n the drai nage of the Rio Cayapas (0 40 ' - 1°15 1N, 78 45 '-79°1 0 1W) i n 1908 . 169 . Jivaro . The Jivaro proper (2°-4°S , 77° - 79°W) i n 1920 . 170 . Amahuaca . The Amahuaca on t he upper I nuya River (10°10 1~ 10°30 1 S , 72°-72°30 ' W) i n 1960 . 171 . I nca . The Quechua- speaking I ndians in the vicinity of Cuz co ( 1 3° 30 r S , 72° W) in 1 530 , i mmediately prior to the Spanish Conquest . 172 . Aymara . The Aymara of the community of Chucui t o in Peru (1 6°S , 70°W) in 1940 . 173 . Si riano . The Siriano in the fore sts near the Rio Blanco (1 4° -1 5°S , 63° - 64°W) in 1942 . 174 . Nambicuara . The Cocozu or eastern Nambicuara (12° 30'- 1303ors, 58° 30 '-59°W) in 1940 . 175 . Trumai . The single survi ving Truma i village (11°50'3 , 53°40 ' W) i n 1938 ~ 176 . Ti mbira . The Ramcocamecra or Eastern Timbira ( 6° - 7°S , 45° - 46°W) i n 1915 . 177 . Tupinamba . The Tupinamba near Rio de Janeiro ( 22°30 1 - 23°S, 42° - 44°30 1W) in 1550 . 178 . Botocudo . The Naknenuk subtribe i n the basin of the Rio Do ce (1 8° - 20°S , 41° 30 '-43°30 ' W) i n 1884 . 179 . Shavante . The Akwe- Shavante in t he vicini ty of Sao Domingos (1 3°30 1 S , 51°30 1 W) i n 1958 . 450 180 . A1-1eikoma . The Aweikoma of the Duque de Caxias Reserva- tion (38°S , 50°W ) in 1932 . 181 . Cayua . The Cayua of southern Mato Grosso , Brazil (23° - 24°S , 54° - 5G 0 W) i n 1890 , the approxi mate period of t he earlier good descriptions~ 182 . Len&ua . The Lengua in contact with the Anglican miss ion (23 - 24°S, 58°-59°W) i n 1889 , t he date of the founding of the mission . 183 . Abi~on . The Abipon i n contact with the Jesuit mission (27 - 29°S , 59° - 66°~) i n 1750 . 184 . Mapuche . The Mapuche i n the vicinity of Temuco (38°30 1 S, 72°35 1 W) in 1950. 185 . Tehue lche or Pata&on . The eouestrian Tehuelche (40° -50°S, 64°-72 W) in 1870: 186 . Yahgan . The eastern and centra l Yahgan (54°30'-55°30 ' 3 , · 67° - 70°W) , reconstructed for 186 5. 45 1 APPENDIX B CODE SHEET FOR ODYSSEY Variable numbers are in parentheses . Where l umping occurs in the format ion of dichot omi es , the 2 categorie s are i ndi cated i n italics . Numbers assigned to t he attributes are those l isted in Appendi x C and are ·the co di ngs for each soci ety fo r each vari able . ( 1) Regional Identification A Afri ca C Circum- Mediterranean E East Eurasia I I nsul ar Pacific N North Ameri ca S South Ameri ca (2 ) Polygyny Absent 1 I ndependent nuclear famili es with monogamy . 2 I ndependent nuclear famil i es with an i nci dence of polygyny of les s t han 20 per cent where permitted . Present 3 Polygyny with an i ncidence of 20 per cent or hi gher . 4 (Varia tions in the Ethnographic Atlas of codi ngs fo r 5 the presence of pol ygyny in sororal , non- sororal , and 6 residence modes are not used . ) (3) Marital Residence I Patrilocal , virilocal 3 Matrilocal , avunculocal, uxori l ocal 452 (4) Marital Residence II Unilocal Patrilocal , virilocal 3 Matri local , avunculo ca l , uxorilocal Other 2 Ambilocal, neolocal (5) Community Organizat ion Endogamous or agamous communit ies 1 Demes (not segmented i nto clan- barrios ) 2 Segmented communities (contai ning locali zed clans ) without local exogamy 3 Agamous communities Exogamous communit ie s 4 Exogamous communit ies (not clans) 5 Segmented communities (contai ning locali zed clans ) 6 Cl an-communi ties (o r clan-barrios ) (6) Settlement Pattern I Nomadic or semisedentary communit ies 1 Fully migratory or nomadic bands 2 Seminomadic communities 3 Semisedentary communities Other forms Variable 7 Variable 8 (7) Settlement Pattern II Settlements , homesteads . and hamlets 4 Compact but i mpermanent settlement s 5 Neighborhoods of dispersed family homesteads 6 Separated hamlets , formlng a sj_ngl e comrnuni ty Other formu Variable 6 Variable 8 453 (8) Settlement Pattern III Compact and complex sett lements 7 Compact and relatively permanent settlements 8 Complex settlements Other forms Variable 6 Variable 7 ( 9 ) Mean Size of Lo cal Communi ties I Fewer than 50 persons 1 Fewer than 50 persons Other mean sizes 2 50-99 persons 3 100-199 persons 4 200- 399 persons 5 400-1 ,000 persons 6 1,000 wi thout any town of more than 5 , 000 persons 7 One or more towns of 5 , 000- 50 , 000 persons 8 One or more cities of more than 50 , 000 ' 10 ) Mean Size of Local Communi t ie s II 50-99 persons 2 50- 99 persons Other mean sizes 1 Fewer than 50 persons 3 100-199 persons 4 200- 399 persons 5 400-1 ,000 persons 6 1, 000 witho~t any town of more than 5 ,000 persons 7 One or more towns of 5 , 000- 50 , 000 persons 8 One or more cities of more than 50 , 000 (1 1) Mean Size of Local Communities III 100-399 persons 3 100-199 persons 4 200-399 persons 454 Other mea n s ize s 1 Fewe r t han 50 persons 2 50- 99 pers ons 5 400- 5 ,000 pers ons 6 1 , 000 wi thou_t any town of more t han 5 ,000 persor,.s 7 One or more towns of 5 ,000- 50 , 000 pers ons 8 One or more cities of more than 50 ,000 (1 2) Local Jurisdictional Hierarchy 0-1 level 1 No leve l s 2 1 level 2-4 levels 3 2 leve ls 4 3 leve ls 5 4 levels (13 ) Jurisdictional Hierarchy Beyond the Local Communi ty I 2- 4 leve l s 3 2 levels--petty paramount chiefdom 4 3 levels- - small state 5 3 leve l s --large state 0- 1 level 1 No levels 2 1 level (14) Juri sdictiona l Hierarchy Beyond the Loca l Community II 1-4 leve l s 2 1 level 3 2 levels--petty paramount chiefdom 4 3 levels--small state 5 4 l evels--la rge state 0 l evels 1 No l evels (1 5) Domina nt Subs istence Activi ty Food-coll ectors 1 Gathering 2 Fi shing 3 Hunting 455 Food- producers 4 Pastoralism 5 I ncipient agriculture 6 Extensive agriculture 7 I ntensive agri cul ture ( 1 6 ) De s cent I 1 Patrilineal 3 Matrilineal ( 17) Descent II Uni l ineal Patrilinea l 3 Matrili neal Non- unil i neal 2 Duolateral 4 Quas i - lineages 5 Ambilineal 6 Bilate ral ( 18 ) Division of Lab or Patridorr.i nant Males alone or a l most a l one 2 Male s appreciably more Non- pat ridoffii nant 3 Di fferent i a tion but equal participation 4 Equal participation without marked differentiation 5 Female s appre ciably more 6 Femal es alone or a lmost a lone 7 Sex participation i rrelevant (1 9 ) Class Stratificat ion Absent 1 Absence among freemen 2 Wealth di st i nctions Present 3 Elite (based on control of land or other re s ource s ) 4 Dual (hereditary aristo cracy ) 5 Complex ( socia l classes ) 456 ( 20) Succession to the Office of Local Headman Here ditary 1 Here ditary patrilineally 2 Hereditary matrilineally Nonhereditary 3 Nonhereditary (appo i nted ) 4 Nonhereditary (seni ority) 5 Nonhereditary (weal th or socia l status ) 6 Nonhereditary (election) 7 Nonhereditary (informal consensus ) ( 21) Freque ncy of I nternal War Freq uent 1 Continual (perpetual , constant) 2 Frequent (common , i ntens i ve ) I nfreq uent 3 I nfrequent or never (occasi onal, s poradic, rare) (2 2) Frequency of External War- Attacki ng Fre quent Cont inual (perpetual , const ant ) 2 Frequent (common , i ntensi ve ) I nfrequent 3 I nfrequent or never ( oc casional , sp oradic , r are) ( 23 ) Frequency of External War- Be i ng At tacked Fre auent 1 Continual (perpetual, const ant ) 2 Frequent ( common , i ntensi ve ) I nfrequent 3 I nfreque• nt or never ( occasi onal , sporadic , rare) ( 24 ) Form of .Military Mobilizat ion 1 Age - grades , military societies , standi ng armie s 2 Men are not organi zed i nt o any of the above ; i nc ludes fri ends , family , lineage, clan ( ki nship organi zat i on and i nforwal non- kinshi p) 457 (25) Decision to Engage i n War 1 1m off icial or council of the polit ical community 2 Anyone (2 6 ) Commencement of War 1 By announcement or mutual agreement 2 By surprise attack (27) Conclusion of War 1 By negotiation 2 By simply st opping , or war is not ended but is con- tinual (28) Peace Ceremony 1 Present 2 .h.bsent (29) Military Expectations I High 1 Subjugation of territory and people 2 Collec t ion of tribute 3 Land--fields , hunting territories , fi shing terri- tories , pastures 5 Trophie s and honors (includes capt i ves for sacrifice ) Low 4 Plunder (includes capt i ves for slaves, hostages , and adopt i on ) 6 Revenge 7 Defense 8 Aggressive defense ( 30) Military Expectations II High 1 Subjugation of t erritory and people 2 Collection of tribute 3 Land--fields , hunting territories, fishi ng terri- tories , pastures 458 Low 4 Plunder (incl udes captives for slaves , hostages , and adoption) 5 Trophies and honors (includes capt ives for sacrifice) 6 Re venge 7 Defense 8 Aggressive defense (31 ) Casualty Rat e 1 High (1/3 or more of combatants ) 2 Low (32) Leadership During Battle 1 An officia l who could back u p his decision by force 2 An i nformal leader whom people obeyed because of res pe ct, but who had no means to force warriors to obey (33) Pr estige Associated with Being A Soldier or Warrior Hi gll 1 A great deal; i mportant for every mal e Low 2 Some ; no t nece ssary to be a warrior to have i nfluence i n t he community 3 No spe cial considerat i on , respe ct, or dist inctions for a man who fights ( 34) Cowardice 1 Refusing to fight (Qefined as ) 2 Leaving companions and the battlefield ; running away ( 35 ) Rewards ( spe cial gift s , praises , or ceremonies , not i ncluding r i tual puri fication ) for a man who has killed an enemy in battle or otherwise shown s kill i n war ) Present/Ela borate 1 Ye s--usually or a lwa ys Absent/not elaborat e 2 Sometimes 3 Rarely or never 459 ( 36 ) Di d members of t he society expect violence to solve t he ir problems ? 1 Ye s 2 No 3 Not clear (37 ) Value of War : Violence/war agai nst non- members of the group i s : Hi gh 1 Enjoyed and considered to have hi gh va l ue Low 2 Cons i dered to be a ne cessary evil 3 Cons i stent l y avo i ded , denounced , no t engaged in (;i8) JVIi l itary Success I : Is the poli t i cal community/cul tural uni t wi nning or l osing i n the long run? Present 1 Yes- - i ts boundari es/ population are expanding Absent 2 No change--boundar i e s /popul ation a r e stati onary (the popul ation i s able t o repl ace t hose l ost i n wa r ) 3 Breaking even--what i t l oses i n territory i t takes from others 4 No--its boundaries/population are shri nki ng ( 39) Military Success II Present Expandi ng 2 No change 3 Breaki ng even Absent 4 Shrinki ng ( 40 ) Subjugation of Terri tory and People 1 Present 2 Absent 460 (41) Collection of Tribute 1 Present 2 Absent (4 2) Land--fields, hunting territories , fishing territories , pastures 1 Pre sent 2 Ab sent (43) Plunder (includes captives for slaves , hostages , adoption) 1 Present 2 Absent (44) Trophies and Honors ( includes captives for sacrifice ) 1 Present 2 Absent ( 4 5) Revenge 1 Present 2 Ab sent (46) Defense Pre sent 2 Absent (47) Aggressive Defense ( defending oneself by attacking first , if it is thought that an enemy is planning to attack) Present 2 Absent Population Density : 1 Less than 'I person per 5 square mi les 2 From 1 person per square mile to 1 per 5 square miles 3 From 1 . 1 to 5 persons per square mile 4 From 5 . 1 to 25 per "'qt1are mj_le 5 From 26 to 100 persons per square mile 6 From 101 to 500 persons per square mile 7 Ove r 500 persons per square mile 46 1 · APPENDIX C . CO DI NG OF VARI ABLES FOR THE STANDARD SAMPLE Variable 1 2 3 4 5 6 7 8 9 10 1 1 1 Nama Hottentot A 2 1 1 6 1 1 1 3 3 3 2 Kung Bushmen A 2 2 2 4 1 1 1 1 1 1 3 Thonga A 4 1 1 6 7 7 7 1 1 1 4 Lo z i A 5 1 1 3 3 3 3 5 5 5 5 Mbundu A 5 1 1 6 7 7 7 4 4 4 6 Kongo A 2 3 3 6 7 7 7 7 Ila A 5 1 1 3 3 3 3 4 4 4 8 Nyakyusa A 5 2 2 4 7 7 7 3 3 3 9 Hadza A 2 1 1 - 1 1 1 1 1 1 10 Luguru A 2 3 3 2 6 6 6 4 4 4 11 Kikuyu A 5 1 1 6 7 7 7 3 3 3 12 Ganda A 5 2 2 4 7 7 7 3 3 3 1 3 Mbut i Pygmie s A 2 2 2 3 1 1 1 2 2 2 14 Nkundo Mango A 5 1 1 6 6 6 6 3 3 3 1 5 Banen A 5 1 1 6 5 5 5 4 4 4 16 Tiv A 5 1 1 6 5 5 5 5 5 5 17 Ibo A 5 1 1 2 6 6 6 - 6 6 " 6 18 Fon A 5 1 1 2 7 7 7 7 7 7 19 Ashant i A 5 3 3 3 7 7 7 4 4 4 20 Mend e A 5 1 1 2 8 8 8 3 3 3 21 Wolof C 5 1 1 6 7 7 7 3 3 3 22 Bambara A 5 1 1 2 7 7 7 5 5 5 23 Tallensi A 5 1 1 6 5 5 5 5 5 5 21~ Songhai C 2 1 1 3 7 7 7 5 5 5 25 Fu l ani C 5 1 1 1 1 1 1 3 3 3 26 Hausa C 5 1 1 2 7 7 7 6 6 6 27 Kanuri C 5 1 1 3 7, 7 7 7 7 7 28 Azande A 5 1 1 3 5 5 5 3 3 3 29 :B'ur C 2 - - 4 7 7 7 3 3 3 30 Otoro Nuba A 5 1 1 2 6 6 6 5 5 5 31 Sbilluk A 5 1 1 2 6 6 6 5 5 5 32 I ngassana A 1 2 2 1 5 5 5 33 Kafa C 6 1 1 - 5 5 5 2 2 2 34 Masai A 5 1 1 3 1 1 1 4 4 4 3 5 Konso C 5 1 1 3 7 7 7 6 6 6 36 Somali C 5 1 1 6 1 1 1 5 5 5 37 Amhara C 1 1 1 4 6 6 6 4 4 4 38 Bisharin C 2 1 1 6 2 2 2 39 Nubi ans C 2 1 1 3 7 7 7 2 2 2 40 Te da C 5 1 1 4 2 2 2 2 2 2 4 1 Tuareg C 1 2 2 6 - - - 1 1 1 42 Riffians C 2 1 1 6 7 7 7 5 5 5 43 Anc . Egypt i ans C 1 2 2 - 7 7 7 8 8 8 44 Hebrews C 6 1 1 - 7 7 7 7 7 7 45 Babyloni ans C 1 2 2 - 7 7 7 8 8 8 46 Rwala Bedouins C 5 1 1 3 1 1 1 2 2 2 47 Turks C 2 1 1 1 7 7 7 5 5 5 48 Gheg Al banians C 6 1 1 4 5 5 5 3 3 3 462 Variab le 1 2 1 3 14 1 5 1 6 1 7 18 1 9 20 21 22 1 Nama Hottentot 3 2 2 3 1 1 1 2 1 2 2 2 Kung Bushmen 4 1 1 1 6 6 5 1 1 3 3 3 Thonga 3 3 3 6 1 1 5 4 1 3 1 4 Lozi 3 4 4 7 5 5 5 4 1 3 2 5 Mbundu 4 3 3 6 2 2 5 4 1 3 2 6 Kongo 4 2 2 6 3 3 5 4 2 2 3 7 Ila 4 2 2 6 2 2 ' 5 2 2 1 1 8 Nyakyusa 3 3 3 7 1 1 2 1 6 2 3 9 Hadza 3 1 1 3 6 6 1 1 4 10 Luguru 4 1 1 6 3 3 2 1 1 3 3 11 Ki kuyu 4 1 1 7 1 1 5 2 2 3 2 12 Ganda 3 4 4 7 1 1 5 4 3 3 2 13 Mbut i Pygmies 3 1 1 1 6 6 5 1 - 3 3 14 Nkundo Mongo 5 3 3 6 1 1 5 1 1 3 1 5 Ban en 3 1 1 6 1 1 5 1 1 2 2 16 Tiv 5 2 2 6 1 1 3 1 - 2 2 17 I bo 5 2 2 6 1 1 5 2 - 2 18 Fon 4 4 4 6 1 1 5 4 1 3 1 19 Ashanti 5 3 3 6 2 2 3 4 6 2 1 20 :i\Tende 5 2 2 6 1 1 3 4 1 2 2 21 Wolof 4 3 3 6 2 2 2 5 1 2 2 22 Bambara 5 2 2 7 1 1 2 4 1 3 2 23 Ta l l ensi 5 1 1 7 1 1 2 2 4 2 3 24 Songhai 4 4 4 7 1 1 2 5 1 - 3 25 Fu lan i 4 2 2 4 1 1 2 1 1 3 2 26 Hausa 4 4 4 7 1 1 2 5 1 1 1 27 Kanuri 4 4 4 7 6 6 4 4 - 2 2 28 Azande 4 3 3 6 1 1 5 4 1 1 1 29 Fur 4 4 4 7 3 3 4 4 1 3 2 30 0toro Nuba 4 2 2 7 1 1 4 2 - 1 2 31 Shilluk 4 2 2 6 1 1 4 5 6 2 1 32 I ngassana 4 1 1 7 5 5 - 1 1 3 2 33 Ka.fa 4 5 5 7 1 1 - 4 34 Masai 4 2 2 4 1 1 3 1 - 2 2 35 Kons o 4 2 2 7 1 1 4 1 1 2 36 Somali 4 3 3 4 1 1 3 2 6 1 2 37 Amhara 4 4 4 7 6 6 2 5 3 3 2 38 Bisharin 4 2 2 4 1 1 1 1 - - 2 39 Nubians 4 1 1 7 1 1 1 2 - 3 3 40 Teda 4 2 2 7 1 1 5 4 4 1 1 41 Tuareg 4 3 3 4 3 3 3 3 2 2 1 42 Riffians 4 2 2 7 1 1 1 2 - 1 2 43 An c. Egyptians 3 5 5 7 6 6 2 5 1 3 1 44 Hebrevrs 4 3 3 7 1 1 1 4 - 3 1 45 Babylonians 3 4 4 7 6 6 1 5 - 3 1 46 Rwala Bedouins 4 2 2 4 1 1 1 2 - 3 1 47 Turks 3 4 4 7 6 6 2 5 6 3 3 48 Gheg Albanians 4 3 3 7 1 1 4 1 1 1 3 463 Variable 23 24 25 26 27 28 29 30 31 32 33 1 Nama Hottentot 2 2 1 3 2 1 1 1 ·2 2 1 2 Kung Bushmen 3 3 - - - - 2 2 2 - 3 3 Thonga 3 1 1 3 - - 2 2 - 1 1 4 Lozi 3 1 1 - - - 1 1 - 1 5 Mbundu - 2 1 - - - 1 1 - 1 6 Kongo 3 2 1 1 1 1 2 2 2 1 2 7 Ila 1 2 2 2 3 - 1 2 1 2 1 8 Nyakyusa 2 1 2 3 2 2 1 1 2 2 1 9 Hadza 2 10 Luguru 2 3 - - - - 2 2 11 Ki kuyu 2 1 1 3 2 1 2 2 2 2 1 12 Ganda 2 1 1 3 2 2 1 1 - 1 1 13 Mbut i Pygmi es 3 3 - - - - - - - - 3 14 Nkundo Mongo 2 2 2 2 1 1 2 2 2 2 15 Banen 2 2 1 - 1 1 1 1 2 - 2 Tiv 2 2 - 3 1 1 1 1 2 2 3 17 Ibo 2 1 1 3 1 1 1 2 · 2 2 2 18 Fon 1 1 . 1 3 2 - 1 1 1 1 1 19 Ashant i 3 1 1 3 1 - 1 2 2 1 1 20 Mende - 1 1 3 2 - 1 1 - 1 1 21 Wolof 1 1 1 3 1 - 1 1 1 - 2 22 Bambara 2 2 1 1 2 - 1 1 - 1 1 23 Tallensi 3 2 2 2 1 - 2 2 2 3 3 24 Songhai 2 1 1 3 1 - 2 2 25 Ful ani 1 2 - - - - 1 1 - 2 2 26 Hausa 3 1 1 3 - - 1 1 - 1 2 27 Kanuri 3 1 1 1 2 - 1 1 2 1 2 28 Azande 2 1 1 3 2 2 1 1 1 1 1 29 Fur 2 2 1 3 1 1 2 2 1 1 2 30 Otoro Nuba 2 1 2 3 1 1 1 2 - 2 1 31 Shilluk 2 2 1 3 1 - 2 2 - 2 1 32 I ngassana - 2 - 3 - 1 2 2 - 1 33 Kafa 34 Masai 2 1 1 3 1 - 2 2 2 2 1 35 Konso 2 1 1 1 2 1 1 1 - - 1 36 Somali 2 2 1 3 1 1 1 1 - 2 1 37 Amhara 3 1 1 1 2 - 1 1 - 2 1 38 Bishari n 2 2 - 3 - - 1 1 - - 1 39 Nub i ans 3 3 - - - - - - - - 3 40 Teda 1 2 2 3 2 - 1 1 - 3 1 4 1 Tuareg 2 1 1 3 2 - 1 1 - 2 1 42 Riffians 2 2 2 3 3 - 1 1 1 3 1 43 Anc . Egypt i ans 2 1 1 - 2 - 1 1 1 1 2 44 Hebrews 3 1 1 1 2 2 1 1 1 1 1 45 Babyloni ans 1 1 1 - 2 - 1 1 - 1 2 46 Rwala Bedouins 1 2 1 1 2 1 1 1 - 2 1 4 7 '.f.'urks 3 1 1 1 1 2 1 1 - 1 1 48 Gheg Albanians 2 2 2 3 3 1 2 2 1 - 1 464 Vari abl e 34 35 36 37 38 39 40 4 1 42 43 44 1 Nama Hottentot - 1 2 2 4 4 2 2 1 1 2 2 Kung Bushmen - 3 2 3 4 4 2 2 2 2 2 3 Thonga 1 1 2 1 4 4 2 2 2 1 2 4 Lozi - - - - 1 1 1 1 1 1 2 5 Mbundu - - - 2 1 1 2 1 2 1 1 6 Kongo - - 2 2 2 2 2 2 2 2 2 7 Ila 1 1 1 1 4 4 2 2 2 1 1 8 Nyakyusa 2 2 2 1 1 1 1 2 1 1 2 9 Hadza - - - - 2 2 10 Luguru - - 2 3 4 4 2 2 2 1 2 11 Ki kuyu 2 1 2 1 1 1 2 2 2 1 2 12 Ganda 2 1 - 1 1 1 2 2 1 1 1 13 Mbut i Pygmies - 3 2 3 4 4 2 2 2 2 2 14 Nkundo Mango - - 2 2 4 4 2 2 2 1 2 15 Banen - 1 2 2 4 4 2 2 1 2 2 16 Tiv - - 1 2 1 1 2 2 1 1 1 17 Ibo - 1 1 1 4 4 2 2 2 1 1 18 Fon - 1 1 1 1 1 1 1 2 1 1 19 Ashanti 2 - 1 1 1 1 2 2 2 1 1 20 :Mende 2 1 1 1 2 2 2 2 1 1 2 21 Wo l of 2 1 1 1 4 4 1 2 2 1 2 22 Bambara 2 1 1 1 1 1 2 2 1 1 2 23 Tallensi 2 3 2 2 2 2 2 2 2 2 2 24 Songhai - - 1 1 4 4 2 2 2 2 2 25 Fulani - - 2 - 4 4 2 2 1 1 1 26 Hausa - 1 1 1 4 4 1 1 1 1 2 27 Kanur j_ - 3 1 1 1 1 1 1 2 1 2 28 Azande 2 1 1 1 1 1 1 1 1 1 1 29 Fur 1 3 1 1 2 2 2 2 2 1 2 30 Ot oro Nuba 1 1 - 1 2 2 2 2 2 1 1 31 Shilluk 2 1 - - 4 4 2 2 2 1 2 32 I ngassana - - - - 2 2 2 2 1 2 1 33 Kafa - - - - - - - - - ....,. 34 Masai 1 1 1 1 4 4 2 2 2 1 2 35 Kons o 3 1 2 2 2 2 2 2 1 1 1 36 Somali - - 1 1 1 1 2 2 1 1 1 37 Amhara 2 1 1 1 1 1 1 2 2 2 2 38 Bi sharin - - 2 2 2 2 2 2 1 1 2 39 Nubians - - 2 3 2 2 40 Teda 1 1 1 1 4 4 2 1 1 1 2 41 Tuareg 1 - 1 1 1 1 2 1 1 1 2 42 Riffians 1 1 1 1 2 2 2 2 1 1 2 43 Anc . Egypt ians 1 1 1 1 1 1 1 1 1 1 2 44 Hebrews 2 3 1 1 1 1 1 2 1 1 2 45 Babylonians - - - 1 4 4 1 2 1 2 2 46 Rwala Bedouins 2 - 1 1 1 1 2 2 1 1 1 4 7 '.l'urks 2 1 - 2 2 2 1 2 2 2 2 48 Gheg Albanians 1 1 1 1 2 2 2 2 2 1 2 465 Variable Population 45 46 47 Densi ty 1 Nama Hottento t 1 1 2 1 2 Kung Bushmen 2 1 2 1 3 T±10nga 2 1 2 6 4 Lozi 2 1 2 4 5 Mbundu 1 2 2 4 6 Kpngo 1 1 2 7 Ila 2 1 2 8 Nyakyusa 1 2 2 5 9 Hadza 2 10 Luguru 1 1 2 7 11 Ki kuyu 2 1 2 6 1 2 Ganda 2 2 1 6 1 3 Mbut i Pygmies 2 2 2 2 14 Nkundo Mongo 2 2 2 3 1 5 Banen 1 1 2 4 1 6 Tiv 1 2 2 5 1 7 I bo 1 1 2 7 1 8 Fon 1 1 2 5 1 9 Ashanti 2 2 2 5 20 Mende 2 1 2 5 21 Wolof 2 2 2 6 22 Bambara 2 2 2 4 23 Tall ensi 1 1 2 6 24 Songhai 2 1 2 4 25 Fulani 2 2 3 26 Hausa 2 ·1 2 5 27 Kanuri 2 2 2 28 Azande 1 1 2 3 29 Fur 2 1 2 4 30 Otoro Nuba 1 2 2 4 31 Shilluk 1 1 2 5 32 I ngassana 2 2 2 33 Kafa 5 34 Masai 2 2 2 3 35 Konso 1 1 2 6 36 Somali 1 2 2 4 37 Amhara 2 2 2 5 38 Bishari n 2 1 2 39 Nubi ans 40 Teda 1 1 2 1 4 1 Tuareg 1 1 2 1 4 2 Riffians 1 1 2 6 43 Anc . Egyptians 2 2 1 44 Hebrews 1 2 2 7 4 5 Babyloni ans 2 1 1 7 46 Hwala Bedouins 1 1 2 2 47 Turks ,, 1 1 c.. 5 48 Gheg Albanians 1 2 2 5 466 Variable 1 2 3 4 5 6 7 8 9 10 1 1 49 Romans C 1 2 2 3 7 7 7 8 8 8 50 Basques C 1 2 2 3 5 5 5 6 6 6 51 I rish C 1 1 1 3 5 5 5 5 5 5 52 La_pps C 1 1 1 4 1 1 1 1 1 1 53 Yurak Samoyed E 2 1 1 4 1 1. 1 1 1 1 54 Russians C 1 2 2 3 8 8 8 6 6 6 55 Abkhaz C 2 1 1 4 7 7 7 4 4 4 56 Armenians C 1 1 1 3 7 7 7 5 5 5 57 Kurd C 2 1 1 1 3 3 3 6 ' 6 6 58 Basseri E 2 1 1 1 1 1 1 3 3 3 59 Punjabi E 1 1 1 4 7 7 7 4 4 4 60 Gond E 3 1 1 6 7 7 7 2 2 2 61 Toda E - 1 1 6 3 3 3 1 1 1 62 Santal E 2 1 1 4 7 7 7 3 3 3 63 Uttar Pradesh E 2 1 1 4 7 7 7 . 6 6 6 64 Burusho E 2 1 1 3 7 7 7 3 3 3 65 Kazak E 6 1 1 6 2 2 2 5 5 5 66 Khalka Mongols E 1 1 1 3 2 2 2 1 1 1 67 Lolo E 2 1 1 6 7 7 7 3 3 3 68 Lepcha E 2 1 1 4 5 5 5 3 3 3 69 Garo E 2 3 3 3 7 7 7 4 4 4 70 Serna Naga E 2 1 ·1 2 7 7 7 4 4 4 71 Burmese E 1 3 3 1 7 7 7 5 5 5 72 Palaung E 2 1 1 1 7 7 7 2 2 2 73 Vietnamese E 2 1 1 2 7 7 7 5 5 5 74 Rhade E 1 3 3 2 6 6 6 5 5 5 75 Kr1mer E 2 2 2 2 8 8 8 7 7 7 76 Siamese E 2 3 3 3 7 7 7 6 6 6 77 Semai E 2 2 2 3 4 4 4 2 2 2 78 Nicobarese E 2 3 3 3 6 6 6 4 4 4 79 Andamanese E 1 2 2 3 2 2 2 1 1 1 80 Vedda E 1 3 3 6 2 2 2 1 1 1 8 1 Tanal a E 2 1 1 5 7 7 7 4 4 4 82 Nebr i Sembilan E 2 3 3 2 7 7 7 6 6 6 83 Javanese I 2 3 3 3 7 7 7 5 5 5 84 Balinese I 1 1 1 1 7 7 7 5 5 5 85 Iban I 1 2 2 3 7 7 7 2 2 2 86 Badjau I 2 2 2 3 1 1 1 4 4 4 87 Toradja I 2 3 3 3 7 7 7 3 3 3 88 Tobelorese I 1 1 1 4 7 7 7 89 Alorese I 2 1 1 4 8 8 8 3 3 3 90 Tiwi I 6 1 1 3 1 1 1 3 3 3 91 Aranda I 5 1 1 6 1 1 1 1 1 1 92 0rokaiva I 2 1 1 2 7 7 7 2 2 2 93 Kimam I 2 1 1 2 5 5 5 5 5 5 94 Kapauku I 6 1 1 6 7 7 7 3 3 3 95 Kwoma I 4 1 1 6 6 6 6 4 4 4 96 Manus I 2 1 1 2 7 7 7 4 4 4 467 Variable 1 2 1) 1 4 1 5 1 6 1 7 18 1 9 20 21 22 49 Romans - - - 7 6 6 - 5 - 3 1 50 Basques - - - 7 6 6 1 5 - 3 3 51 I r i sh 3 3 3 7 6 6 2 5 - 3 2 52 Lapps 3 1 1 4 6 6 4 2 - 3 3 53 Yu.rok Samoyed 4 1 1 3 1 1 1 2 - - 2 54 Russ i ans 5 5 5 7 6 6 3 1 3 3 2 55 Abkhaz - - - 4 1 1 2 - 4 - 2 56 Armenians 4 4 4 7 6 6 2 2 6 3 57 Kurd 4 3 3 7 1 1 2 4 1 2 2 58 Bas seri 3 2 2 4 1 1 2 2 1 2 2 59 Punjabi 4 5 5 7 1 1 1 - 1 3 2 60 Gond 4 2 2 6 1 1 3 2 1 3 3 61 Toda 4 2 2 4 2 2 1 1 7 3 3 62 Santal 4 2 2 7 1 1 4 1 1 3 3 63 Uttar Pradesh 4 5 5 7 1 1 1 - 5 3 2 64 Burusho 5 2 2 7 1 1 4 5 - - 2 65 Kazak 4 4 4 4 1 1 2 4 - 2 2 66 Khal ka Mongol s 4 4 4 4 1 1 3 4 1 2 2 67 Lolo 4 1 1 7 1 1 4 4 1 1 2 68 Lepcha 4 2 2 7 1 1 4 2 1 3 3 69 Garo 4 2 2 6 3 3 3 2 2 1 2 70 Serna Naga 4 3 3 6 1 1 4 4 - 2 2 71 Burme se 3 4 4 7 6 6 2 5 3 2 3 72 Palaung 3 1 1 6 6 6 4 2 1 3 3 73 Vietnamese 4 5 5 7 1 1 2 5 6 2 1 74 Rhade 4 1 1 6 3 3 2 2 2 - 3 75 Khmer 5 5 5 7 6 6 4 - 1 3 1 76 Si amese 4 5 5 7 6 6 4 5 6 2 2 77 Semai 4 1 1 6 6 6 3 1 - 3 3 78 Nicobarese 4 1 1 6 6 6 2 1 7 3 3 79 Andamanese 3 1 1 2 6 6 3 1 7 2 3 80 Vedda 4 1 1 1 3 3 3 1 - 3 3 81 Tanala 5 2 2 7 1 1 3 4 6 3 3 82 Negri Sernbi l an 5 5 5 7 3 3 - - 6 2 3 83 Javanese 3 5 5 7 6 6 5 5 6 3 2 84 Balinese 4 3 3 7 5 5 1 4 1 2 85 I ban 4 1 1 6 6 6 5 1 7 2 2 86 Badjau 3 1 1 · 2 6 6 2 2 1 3 3 87 Torad ja 4 2 2 6 6 6 4 2 6 88 Tobelore se 4 3 3 6 6 6 4 1 6 89 Alorese 4 3 3 6 1 1 5 2 - 3 3 90 Tiwi 4 1 1 1 3 3 6 1 1 2 3 91 Aranda 3 1 1 1 2 2 6 1 1 3 3 92 Orokai va 4 1 1 6 1 1 5 1 7 2 2 93 Ki mam 5 1 1 7 1 1 2 1 - 2 94 Kapauku 4 2 2 6 1 1 3 2 5 2 3 95 Kworna 4 1 1 5 1 1 4 1 - 3 3 96 Manus 4 1 1 2 2 2 2 2 - 3 3 468 Vari able 23 24 25 26 27 28 29 30 31 32 33 49 Romans 2 1 1 1 1 - 1 1 1 1 2 50 Basq_ues 2 -1 1 3 1 2 1 1 - 1 2 51 I rish 3 1 1 3 1 1 1 1 2 2 2 52 Lapps 3 - - - - - 2 2 - - 3 53 Yurak Samoyed - 2 - - - - 1 1 - - 3 54 Russians 3 1 1 3 2 - 1 1 1 1 2 55 Abkhaz 2 1 1 3 - - 1 1 - 1 2 56 Armenians 2 - - - - - 2 2 1 - 3 57 Kurd 2 2 1 1 1 - 1 1 - 1 1 58 Basseri 2 2 1 - 1 - 1 1 - 1 2 59 Punjabi 3 1 1 3 1 2 1 1 - 1 2 60 Gond 3 - - - - - 1 2 2 61 Toda 3 3 - - - - 2 2 - - 3 62 Santal 3 3 1 - 2 - 1 1 1 2 2 63 Uttar Pradesh 2 2 2 3 1 - 1 1 - 2 2 64 Burusho 3 - 1 3 - - 2 2 65 Kazak - 2 2 3 2 - 1 1 2 1 2 66 Khalka Mongol s 2 1 1 - 1 - 2 2 - 1 2 67 Lalo 2 2 1 3 1 1 2 2 2 2 1 68 lepcha 3 3 - -- - - 2 2 - - 3 69 Garo 1 2 2 3 2 - 1 1 - 2 70 Serna Naga 3 2 1 1 1 - 1 1 2 1 1 71 Burmese 2 1 1 3 - - 1 1 1 1 1 72 Palaung 3 1 1 3 - 1 1 1 - - 2 73 Vi etnamese 1 1 1 3 3 2 2 2 1 1 2 74 Rhade 1 - 1 3 - - 2 2 - 1 2 75 Khmer 1 1 1 - 1 1 1 1 1 1 76 Siamese 3 1 1 - 1 - 1 1 - 1 2 77 Semai 2 3 2 - 1 - 2 2 - 3 3 78 Nicobarese 3 3 - - - - 1 1 - - 3 79 Andamanese 3 2 - 3 1 1 2 2 2 3 80 Vedda 3 - - - - - 2 2 - - 3 81 Tanal a 3 1 1 3 2 1 1 1 - 1 2 82 Negri Sembilan 3 1 1 3 2 - 1 1 2 1 2 83 J avanese 1 1 1 3 1 1 1 1 1 1 2 84 Bal inese 3 1 1 - 1 - 1 1 1 1 2 85 I ban 2 2 1 3 1 1 1 1 1 2 1 86 Badjau 3 3 - - . - - 2 2 - - 3 87 Toradja. - 2 2 3 1 1 1 1 2 88 Tobelorese 89 Alorese 3 2 2 3 1 1 2 2 2 3 3 90 Tiwi 3 2 2 2 3 1 2 2 2 3 1 91 Aranda 3 2 1 2 3 - 2 2 2 2 1 92 Orokaiva 2 2 2 3 3 1 1 2 - 2 1 93 Ki mam - 2 - - 1 1 1 2 - - 1 94 Kapauku 3 2 1 1 1 1 2 2 1 2 2 95 Kwoma 3 1 1 3 1 - 1 1 2 3 1 96 Manus - 2 1 3 2 1 1 1 2 2 2 46 9 Variable 34 35 36 37 38 39 40 41 42 43 44 4 9 Romans 1 2 1 1 1 1 1 1 2 1 2 50 Basques - 3 1 2 4 4 1 2 2 2 2 51 Iri sh - 2 2 2 1 1 1 2 2 2 2 52 Lapps - 3 2 3 2 2 2 2 2 2 2 53 Yurak Samoye d - - 2 - 4 4 2 2 1 1 2 54 Russians 1 2 1 1 1 1 1 2 2 2 2 55 Abkhaz 1 - - - 4 4 1 2 2 2 2 56 Armeni ans - - - - 4 4 2 2 2 2 2 57 Kurd - 1 - 1 4 4 1 1 1 1 2 58 Basseri - - 2 2 1 1 1 2 1 2 2 59 Punjabi - - 1 1 2 2 1 2 2 2 1 60 Gond - - 2 - 2 2 2 2 2 2 1 61 Toda - 3 2 3 2 2 2 2 2 2 2 62 Santal - 2 2 2 4 4 2 2 1 2 2 63 Uttar Pradesh - 3 1 2 1 1 1 2 1 2 2 64 Burusho - - 2 1 4 4 2 2 2 1 2 65 Kazak - - 2 1 2 2 2 2 1 1 2 66 Khalka Mongols - - 2 2 4 4 2 2 2 2 2 6 7 Lo l o 1 1 1 1 3 3 2 2 2 1 2 68 Lepcha - 3 2 3 4 4 2 2 2 2 2 69 Garo 1 2 1 - 3 3 2 2 1 1 1 70 Serr..a Naga - 1 - 1 1 1 2 2 1 2 1 71 Burmese - 1 1 2 2 2 1 2 2 2 2 72 ?alaung - - 2 3 2 2 1 2 2 2 2 73 Vie tnamese - 1 1 2 4 4 2 2 2 2 2 74 Rhade - - 2 3 4 4 2 2 2 2 2 75 Khmer - - - 1 1 1 1 2 2 1 2 76 Siamese - - 1 - 2 2 2 2 1 2 2 77 Sema i - - 2 3 4 4 2 2 2 2 2 78 Ni c obare s e - - 2 3 4 4 2 2 1 2 2 79 Andamanese - - - - 4 4 2 2 2 1 2 80 Vedda - 3 2 2 4 4 2 2 2 2 2 8 1 Tanala 1 - 2 2 2 2 1 2 2 1 2 82 Negri Sembilan -- - - 3 2 2 2 1 1 2 2 83 Javanese - - 1 2 1 1 1 2 2 2 2 84 Baline se - - 2 3 2 2 1 2 2 1 2 85 I ban - 1 1 1 1 1 2 2 1 1 1 86 Bad jau - - 2 3 2 2 2 2 2 2 2 8 7 Toradja 1 1 - - - - 2 2 1 1 1 88 Tobelorese 89 Alorese - 1 2 1 2 2 2 2 2 2 2 90 Tiwi - 1 1 2 2 2 2 2 2 2 2 91 Aranda 2 - 1 1 2 2 2 2 2 1 2 92 0rokai va - 1 1 - 1 3 3 2 2 2 1 1 93 Ki mam - - - - 4 4 2 2 2 2 1 94 Kapauku 1 1 2 2 2 2 2 2 2 1 2 95 Kwoma 1 1 1 2 2 2 2 2 1 2 1 96 Manus - 3 2 3 2 2 2 2 1 1 2 Variable 45 46 Populat i on 47 Density 49 Homans 2 1 1 7 50 Basques 2 1 2 4 51 Irish 1 1 2 5 52 Lapps 1 1 2 3 53 Yurak Samoyed 1 1 2 1 54 Russ i ans 2 1 2 4 55 Abkhaz 2 1 2 5 56 Armeni ans 2 1 2 5 57 Kurd 1 2 2 6 58 Basseri 2 2 2 3 59 Punjabi 1 1 2 6 60 Gond 2 2 2 4 61 Toda 2 2 2 5 62 Santal 2 1 2 6 63 Uttar Pradesh 1 2 2 7 64 Burusho 2 2 2 3 65 Kazak 1 2 2 3 66 Khal ka :Mongol s 2 1 2 3 67 Lolo 1 2 2 3 68 Lepcha 2 1 2 5 69 Garo 1 1 2 5 70 Serna Naga 1 2 2 71 Burmese 2 1 2 6 72 Palaung 2 2 2 73 Vietnamese 2 1 2 7 74 Rhade 1 2 2 7 75 Khmer 2 1 2 5 76 Siamese 2 1 2 6 77 Semai 2 2 2 1 78 Nicobare se 2 , 2 2 5 79 Andamanese 1 2 2 3 80 Vedda 1 1 2 1 81 Tanal a 2 1 2 3 82 Negri Sembilan 1 1 2 7 83 Javanese 2 1 1 7 84 Balinese 2 1 2 7 85 Iban 2 1 2 4 86 Badjau 2 2 2 1 87 Toradja 1 1 2 4 88 Tobelore se 4 89 Alorese 1 2 2 7 90 Tiwi 1 1 2 2 91 Aranda 1 1 1 2 92 0rokaiva 1 1 1 3 93 Kimam 1 1 2 2 94 Kapauku 1 1 2 6 95 Kwoma 1 2 2 5 96 :Manus 2 2 5 4 7 1 Variable 1 2 3 4 5 6 7 8 9 10 1 1 97 New Ireland I 2 3 3 2 6 6 6 4 4 4 98 Trobrianders I 2 3 3 2 7 7 7 3 3 3 99 Siuai I 2 3 3 2 6 6 6 2 2 2 100 Tikopia I 5 1 1 3 6 6 6 5 5 5 101 }')entecost I 2 1 1 2 7 7 7 2 2 2 1 0 2 Ifoau Fijians I 2 1 1 4 7 7 7 6 6 6 103 Ajie I 5 1 1 6 7 7 7 2 2 2 104 Maori I 6 1 1 2 7 7 I 7 3 3 3 105 Marquesans I 3 1 1 3 5 5 5 3 3 3 106 Samoans I 2 2 2 2 7 7 7 4 4 4 1 07 Gilbertese I 2 1 1 2 8 8 8 4 4 4 108 Marshallese I 2 2 2 2 7 7 7 3 3 3 109 Trukese I 2 3 3 2 6 6 6 4 4 4 11 0 Yapese I 2 1 1 2 5 5 5 3 3 3 111 Palauans I 2 3 3 2 7 7 7 2 2 2 11 2 Ifugao I 2 2 2 3 6 6 6 5 5 5 11 3 Atayal I 1 1 1 3 7 7 7 3 3 3 11 4 Chinese E 1 1 1 4 7 7 7 6 6 6 1 15 Manchu E 1 1 1 6 7 7 7 3 3 3 11 6 Koreans E 1 1 1 4 7 7 7 3 3 3 117 Japanese E 1 1 1 3 8 8 8 3 3 3 118 Ai nu E 2 1 1 1 3 3 3 1 1 1 11 9 Gilyak E 2 1 1 2 2 2 2 2 2 2 120 Yukaghir E 2 3 3 4 2 2 2 2 2 2 1 21 Chukchee E 5 1 1 3 1 1 1 1 1 1 122 I ngalik N - 3 3 3 2 2 2 2 2 2 123 Aleut N 6 1 1 3 7 7 7 4 4 4 124 Copper Eskimo N - 2 2 1 2 2 2 1 1 1 125 Montagnais N 2 1 1 3 2 2 2 3 3 3 126 Micmac N - 1 1 3 2 2 2 1 1 1 127 North Saulteaux N 2 1 1 3 2 2 2 3 3 3 128 Slave N - 1 1 3 2 2 2 2 2 2 129 Kaska N 1 3 3 4 2 2 2 1 1 1 130 Eyek N - 3 3 3 7 7 7 1 1 1 131 Haida N 2 3 3 6 7 7 7 3 3 3 132 Be llacoola N 6 1 1 3 7 7 7 2 2 2 133 Twana N 2 1 1 4 3 3 3 1 1 1 134 Yurok N 6 1 1 4 7 7 7 1 1 1 1 35 East Pomo N 2 2 2 3 3 3 3 4 4 4 136 Yokuts N 5 1 1 2 2 2 2 137 Paiute N - 3 3 3 2 2 2 3 3 3 138 Klamath N 2 1 1 3 2 2 2 1 1 1 139 Kutenai N - 1 1 3 ') c.. 2 2 3 3 3 140 Gros Ventre N - 1 1 4 1 1 1 3 3 3 141 Hidatsa N 3 3 3 3 3 3 3 5 5 5 142 Pawnee N 3 3 3 1 3 3 3 4 4 4 143 0n1aha N - 1 1 2 3 3 3 6 6 6 144 Huror: N 1 3 3 3 3 3 3 6 6 6 472 Variable 1 2 1 3 14 1 5 1 6 1 7 18 1 9 20 21 22 97 New Ireland 4 1 1 5 3 3 5 1 5 2 98 Trobrianders 3 2 2 6 3 3 3 4 2 2 3 99 Siuai 4 1 1 6 3 3 4 2 5 2 100 Tikopia 4 2 2 5 1 1 6 4 1 3 3 1 01 Pentecost 4 1 1 6 2 2 2 2 2 102 Moau Fijians 5 5 5 2 1 1 1 2 103 Ajie 4 2 2 7 1 1 5 1 1 2 2 104 Maori 5 2 2 5 5 5 2 4 1 1 3 1 05 Marque sans 4 2 2 6 6 · 6 1 4 1 3 2 106 Samoans 5 3 3 6 5 5 3 4 1 2 107 Gilbertese 4 2 2 5 5 5 5 4 4 3 108 Marshallese 4 2 2 6 3 3 2 4 2 2 3 109 Trukese 4 1 1 5 3 3 2 1 2 2 2 1 1 0 Yapese 4 2 2 6 2 2 5 5 2 2 3 1 1 1 Palauans 5 3 3 6 3 3 5 4 2 2 3 11 2 Ifuga o 4 1 1 7 6 6 5 2 2 11 3 Atayal 4 2 2 6 1 1 5 1 . 6 1 14 Chine se 5 5 5 7 1 1 4 7 2 3 11 5 Manchu 5 7 1 1 2 2 6 3 1 1 6 Koreans 4 3 3 7 1 1 4 5 4 3 3 11 7 Japanese 4 5 5 7 6 6 4 5 6 3 1 11 8 Ainu 4 2 2 2 3 3 3 1 1 2 2 11 9 Gilyak 3 1 1 2 1 1 1 1 1 2 3 120 Yukaghir 4 1 1 3 6 6 1 1 4 3 2 1 21 Chukchee 4 1 1 2 6 6 1 2 5 3 2 122 Ingalik 3 1 1 2 6 6 2 2 4 3 3 12 3 Aleut 4 1 1 2 4 4 1 4 1 2 1 124 Copper Eski mo 3 1 1 2 6 6 2 1 3 3 125 Montagnais 4 1 1 3 6 6 1 1 1 3 3 1 26 Micmac 3 2 2 3 6 6 1 1 1 3 2 127 North Saulteaux 4 1 1 3 1 1 1 1 1 3 3 1 28 Slave 3 1 1 3 6 6 1 1 7 3 3 129 Kaska 3 1 1 2 3 3 4 1 7 3 3 130 Eyak 4 2 2 2 3 3 1 2 2 3 2 1 31 Haida 4 1 1 2 3 3 1 4 2 2 1 1 32 Bella coola 4 1 2 5 5 1 4 7 3 3 1 33 Twana 4 1 1 2 6 6 2 2 5 3 3 1 34 Yurok 4 1 1 2 6 6 2 2 5 2 3 1 35 East Porno 4 1 1 1 6 6 5 2 2 3 3 1 36 Yokuts 4 1 1 1 1 1 4 2 1 3 3 137 Paiute 3 1 1 1 6 6 6 1 7 3 3 138 Klamath 4 1 1 2 6 6 2 2 7 3 1 139 Kutenai 3 2 2 2 6 6 1 1 5 3 3 14 0 Gr-os Ventre 3 2 2 3 6 6 1 1 5 1 1 141 Hidatsa 4 1 1 7 3 3 6 1 7 3 2 142 Pawnee 4 2 2 6 3 3 6 4 1 3 2 143 Omaha 4 2 2 5 1 1 5 1 5 3 3 144 Huron 4 2 2 5 3 3 6 2 2 3 1 473 Variable 23 24 25 26 27 28 29 30 31 32 33 97 New I reland 3 1 2 2 2 2 98 Trobrianders 2 2 2 2 1 2 1 1 99 Siuai 2 2 1 2 1 1 3 2 100 Tikopi a 3 3 2 2 2 3 1 01 Pente cos t 1 2 3 1 2 2 3 2 102 Moau Fijians 2 1 1 2 1 2 1 1 2 103 Ajie 2 2 1 3 1 1 1 1 2 104 Maori 3 2 1 3 2 1 1 1 1 2 1 105 Marq_ue sans 2 2 1 1 2 1 1 2 2 1 2 106 Samoans 2 1 1 2 1 1 1 1 2 107 Gilbert ese 2 1 1 1 1 2 108 Marshalle se 2 2 1 2 1 1 1 2 1 1 109 Trukese 2 2 1 1 1 2 2 2 2 2 11 0 Yapese 3 2 1 1 1 2 2 2 1 2 1 1 1 Pal auans 3 1 ·1 1 1 1 1 11 2 I fuga o 2 2 1 3 3 1 2 2 2 2 1 11 3 Atayal 2 2 11 4 Chi nese 2 1 1 3 1 1 1 2 1 1 5 Manchu 1 1 2 2 1 11 6 Koreans 2 1 1 3 2 2 2 1 2 11 7 Japanese 1 1 1 1 2 1 1 1 1 1 1 11 8 Ai nu 2 3 1 3 1 1 1 1 1 1 2 11 9 Gi lyak 3 2 2 3 1 1 2 2 1 2 3 ·j 20 Yukaghir 2 2 1 3 1 1 1 2 2 1 2 1 Chukchee 3 2 3 2 2 2 2 1 1 22 I nga lik 3 2 1 3 2 2 2 3 3 123 Aleut 2 2 1 3 1 1 1 2 1 2 1 1 24 Coppe r Eskimo 3 3 3 125 Montagnais 2 2 2 2 2 1 2 2 2 126 Micmac 2 2 1 3 2 1 1 2 2 1 127 North Saulteaux 3 3 3 128 Slave 3 2 2 3 1 2 2 2 2 12 9 Kaska 3 1 2 3 2 1 2 2 2 2 130 Eyak 2 2 1 3 2 2 2 2 2 1 2 1 31 Haida 2 1 3 1 1 1 1 1 2 2 1 32 Bellacoola 2 1 2 3 2 1 1 1 2 2 2 133 Twana 3 1 2 2 2 2 2 2 134 Yurok 3 2 2 3 1 1 2 2 1 2 2 1 35 Eas t Pomo 3 2 1 3 1 1 2 1 2 2 2 1 36 Yokuts 3 2 2 2 1 2 2 2 2 3 1 37 Pa i ute 2 3 2 2 2 2 138 Klamath 2 2 2 3 2 2 2 2 2 2 139 Kutenai 2 2 2 3 2 1 1 2 1 140 Gros Ventre 1 2 3 1 2 2 2 1 14 1 Hi datsa c"c . 1 1 3 1 1 1 1 1 2 2 142 Pawnee 1 1 2 3 2 2 1 2 2 2 143 Omaha 2 1 3 1 1 1 1 2 1 144 Huron 2 1 3 1 1 2 2 1 1 474 Variable 34 35 36 '57 38 39 40 41 42 43 44 97 Ne-w Ireland 2 2 2 2 2 2 2 1 2 98 Trobrianders 2 2 2 2 2 1 1 98 Siuai 2 1 2 2 2 1 2 2 2 100 Tikopia 2 2 2 2 2 2 2 1 0 1 Pentecost 2 2 2 2 2 2 2 2 102 Moau Fi jians 4 4 2 2 2 1 1 103 Ajie 4 4 2 2 1 1 1 1 04 Maori 1 1 1 1 1 2 2 1 1 2 1 05 Marque sans 1 1 1 3 3 2 2 2 1 1 106 Samoans 3 1 1 1 3 3 1 2 2 1 1 107 Gi l bertese 2 2 2 2 1 2 2 108 Marshallese 2 1 1 2 2 1 2 2 2 2 109 Trukese 2 2 2 2 2 2 2 2 2 1 1 0 Yapese 2 1 4 4 2 2 2 1 2 1 1 1 Palau.ans 1 2 2 1 1 2 1 2 1 1 2 I f ugao 1 1 2 2 2 2 2 1 2 1 1 3 Atayal · 1 1 - . - 11 4 Chinese 2 2 4 4 1 2 2 2 2 1 1 5 Manchu 1 4 4 2 2 2 2 2 1 1 6 Koreans 2 2 2 4 4 2 2 2 2 2 11 7 Japar..ese 2 1 1 4 4 1 2 1 2 2 118 Ainu 2 4 4 2 · 2 1 1 2 11 9 Gilyak 2 2 2 2 2 2 1 2 120 Yukaghir 2 1 2 4 4 2 1 2 1 2 1 21 Chukchee 2 1 1 1 1 2 2 2 1 2 1 22 Ingalik 3 1 1 2 2 2 2 2 1 2 123 Aleut 2 1 1 1 4 4 2 2 2 1 1 1 24 Copper Eskimo 1 3 2 2 1 25 Montagnais 2 2 1 2 2 2 2 2 1 1 126 Micmac 1 1 1 2 2 2 2 2 1 1 127 North Saulteaux - 3 1 3 1 1 128 Slave 3 2 2 2 2 2 2 2 1 2 129 Kaska 1 2 2 4 4 2 2 2 1 2 130 Eyak 3 1 2 2 2 2 2 2 1 2 1 3 1 Haida 2 2 2 1 1 2 2 1 1 1 1 32 Bella coola 2 2 2 2 2 1 1 1 1 33 Twana 3 2 3 2 2 2 2 2 1 2 1 34 Yurok 2 1 2 2 2 2 2 2 2 2 1 35 East Pomo 2 2 2 2 2 2 2 1 2 1 1 36 Yokuts 4 4 2 2 2 1 2 137 Paiute 3 2 2 4 4 2 2 2 2 2 138 Klamath 1 1 1 2 2 2 2 2 1 2 139 Kutenai 2 1 1 2 2 2 2 1 1 1 140 Gros Ventre 1 1 4 4 2 2 2 1 1 141 Hidatsa 2 1 2 2 4 4 2 2 1 1 1 142 Pawnee 1 1 2 . 1 4 4 2 2 2 1 1 143 Omaha - 1 2 2 3 3 2 2 1 1 1 144 Huron 1 1 1 2 2 2 2 2 2 2 475 Variable Population 45 46 47 Density 97 New I reland 1 2 2 4 98 Trobrianders 2 2 2 5 99 Siuai 1 2 5 100 Tikop i a 2 1 2 6 1 01 Pente cost 1 1 2 4 102 Moau Fi j i ans 1 1 2 7 103 Ajie 2 1 2 3 104 Maori 1 1 2 2 105 Mar que sans 1 2 2 5 106 Samoans 1 1 2 6 1 07 Gilberte se 1 1 2 6 108 M:ars halle se 1 2 2 6 109 Tr uke se 1 2 2 7 11 0 Yape se 1 2 2 5 1 1 1 Pal auans 2 2 2 5 11 2 I fugao 1 2 2 6 11 3 Atayal 1 4 11 4 Chi nese 2 1 2 7 11 5 Manchu 1 1 2 6 11 6 Koreans 2 1 2 7 11 7 Japanese 2 1 2 7 11 8 Ai nu 2 2 2 2 11 9 Gi lyak 1 ·1 2 1 120 Yukaghi r 1 . 2 2 1 1 21 Chukchee 2 2 2 1 122 I ngali k 1 2 2 1 123 Al eut 1 2 2 4 124 Copper Eski mo 1 1 25 Montagnai s 1 2 2 1 1 26 Micmac 1 2 2 1 127 North Saul teaux - 1 128 Slave 1 2 2 2 129 Kaska 1 1 1 1 130 Eyak 1 1 2 1 1 3 1 Hai da 1 1 2 2 132 Bell acoola 1 2 3 1 33 Twana 1 1 2 3 134 Yur ok 1 2 2 3 1 35 East Pomo 1 2 2 3 1 36 Yokuts 1 2 2 4 13 7 Paiute 1 1 2 1 1 38 Klamath 1 1 2 2 139 Kutenaj_ 1 1 2 1 14 0 Gros Ventre 1 1 1 14 1 Hi datsa 1 1 2 1 142 Pawnee 1 1 2 2 143 Omaha 1 1 2 3 144 Huron 1 1 2 5 476 Vari ab l e 1 2 3 4 5 6 7 8 9 10 1 1 145 Creek N 2 3 3 2 7 7 7 4 4 4 146 Nat chez N 3 1 1 3 7 7 7 5 5 5 147 Comanche N 3 2 2 1 1 1 1 4 4 4 148 Apache N 3 3 3 3 1 1 1 2 2 2 149 Zuni N 1 3 3 2 7 7 7 6 6 6 150 Havasupai N 2 1 1 1 3 3 3 3 3 3 151 Papago N 3 1 1 4 3 3 3 4 4 4 152 Hui chol N 2 2 2 1 6 6 6 4 4 4 1 53 Aztec N - 1 1 2 7 7 7 8 8 8 1 54 Popoluca N - 1 1 - 7 7 7 5 5 5 155 Yucate c Maya s 1 1 1 3 8 8 8 7 7 7 1 56 Miski t o s 2 3 3 3 7 7 7 4 4 4 1 57 Bri bri s 2 2 2 3 5 5 5 2 2 2 1 58 Cuna s 2 3 3 1 7 7 7 5 5 5 1 59 Goa jiro s 5 3 3 3 1 1 1 2 2 2 160 Hai t i ans s 2 1 ·1 2 7 7 7 7 7 7 · 1 61 Callinago s 2 3 3 2 6 6 6 . 4 4 4 162 Warrau s 2 3 3 1 3 3 3 2 2 2 163 Yanomamo s 6 1 1 2 4 4 4 3 3 3 1 64 Cari b s 2 2 2 3 5 5 5 1 1 1 165 Saramacca s 5 3 3 2 7 7 7 3 3 3 166 Mu.ndurucu s 2 3 3 3 7 7 7 3 3 3 167 Cubeo s 2 1 1 6 4 4 4 1 1 1 168 Cayapa s 1 1 1 1 5 5 5 5 5 5 169 J i varo s 6 2 2 3 5 5 5 2 2 2 170 Amahuaca s 2 1 1 4 5 5 5 1 1 1 1 7 1 Inca s 1 1 1 1 7 7 7 3 3 3 172 Aymara s 1 1 1 1 7 7 7 5 5 5 173 Siriono s 3 3 3 2 2 2 2 2 2 2 174 Nambicuara s 2 1 1 3 2 2 2 2 2 2 175 Trumai s 2 1 1 3 7 7 7 1 1 1 176 Ti mbira s 1 3 3 2 2 2 2 4 4 4 177 Tupi namba s 5 3 3 3 7 7 7 5 5 5 1 78 Bo t ocudo s 3 - - 4 1 1 1 2 2 2 179 Shavante s 6 3 3 3 2 2 2 4 4 4 180 Awe i koma s 6 2 2 1 1 1 1 1 1 1 181 Cayua s 1 1 1 3 . 7 7 7 1 1 ·1 182 Lengu.a s 2 3 3 3 1 1 1 3 3 3 183 Abipon s 2 2 2 3 1 1 1 4 4 4 184 Mapuche s 5 1 1 5 6 6 6 2 2 2 185 Tehuelche s 2 1 1 3 1 1 1 3 3 3 186 Yahgan s 2 1 1 4 1 1 1 1 1 1 477 Variable 1 2 1 3 14 1 5 1 6 1 7 18 1 9 20 21 22 14 5 Cree k 5 3 3 5 3 3 5 1 2 3 1 146 Nat chez 4 ') c.. 2 5 6 6 5 3 3 3 2 147 Comanche 4 3 6 6 1 7 3 1 148 Apache 4 1 1 6 6 6 7 3 1 149 Zuni 5 1 7 3 3 2 1 2 3 2 1 50 Havasupai 4 1 5 4 4 2 1 7 3 3 1 5 ·1 Papago 4 1 5 6 6 2 1 1 3 2 1 52 Huichol 4 1 6 6 6 2 2 6 3 3 1 53 Aztec 4 3 3 7 5 5 2 5 3 1 1 54 Popo luca 3 1 1 6 6 6 · 1 1 6 3 1 55 Yucatec Maya 4 2 2 6 1 1 1 5 1 1 1 56 Mi ski to 4 2 2 1 6 6 4 1 3 1 1 57 Bribr i 3 1 1 6 3 3 4 2 1 58 Cuna 4 2 2 6 6 6 1 2 6 3 1 59 Goajiro 4 1 1 4 3 3 3 2 2 2 3 160 Haitians 4 1 1 5 6 6 3 1 3 1 61 Callinago 4 1 1 2 3 3 2 1 7 2 162 Warrau 4 1 1 1 6 6 1 7 3 3 163 Yano;namo , 5 1 1 5 1 1 2 1 5 1 2 164 Carib 3 1 1 5 4 4 3 1 7 3 3 165 Saramacca 4 3 3 5 3 3 5 1 2 3 1 1 66 :Mundurucu 4 1 1 5 1 1 3 1 1 3 2 1 67 Cuoeo 3 1 1 5 1 1 5 1 1 2 3 168 Cayapa 4 1 1 6 6 6 5 1 1 3 169 Jivaro 3 1 6 6 6 5 1 1 170 .Amahuaca 3 1 1 5 6 6 3 1 1 71 I nca 3 5 5 7 6 6 2 5 3 2 172 Aymara 4 6 1 1 4 2 2 2 173 Siri ono 4 1 1 3 6 6 1 1 1 3 3 174 Nambicuara 4 1 1 1 6 6 5 1 7 2 2 1 75 Trumai 4 1 1 5 4 4 3 1 1 3 3 176 Timbira 5 1 1 5 3 3 5 1 3 2 177 11upinamba 4 2 2 5 6 6 5 2 1 1 1 178 Botocudo 4 ·1 1 1 6 6 6 1 7 1 2 179 Shavante 5 1 1 1 1 1 5 1 5 2 2 180 Awe ikoma 3 1 1 3 6 6 1 1 1 2 18 1 Cayua 3 1 1 5 6 6 2 1 3 182 Lengua 4 1 1 3 6 6 1 1 1 . 183 Abipon 3 1 1 3 6 6 1 4 1 3 2 184 Mapuche 5 1 1 7 1 1 2 2 1 2 1 185 i'ehuelche 4 1 1 3 6 6 1 1 1 2 2 186 Yahgan 3 1 1 2 6 6 3 1 2 3 478 Variable 23 24 25 26 27 28 29 30 31 32 33 145 Creek 1 2 1 3 1 1 1 2 - 1 1 146 Nat chez 2 1 1 3 2 1 1 1 - 2 2 147 Comanche 1 2 2 3 1 1 1 1 2 2 1 148 Apache 1 2 1 3 3 - 1 1 - 2 1 149 Zuni 2 1 1 3 1 1 2 2 - 2 2 150 Havasupai 3 2 2 3 1 2 1 2 2 2 2 151 Papago 1 2 1 3 2 2 1 2 1 2 1 152 Huichol 3 3 - - -· - - - - - 3 153 Aztec 1 1 1 1 1 2 1 1 - 1 1 154 Popoluca - 1 - - - - - - - - 2 15 5 Yucatec Maya 1 1 1 3 3 - 1 1 - 1 1 156 Miskito 1 1 1 3 3 - 1 1 - 1 1 157 Bribri 3 - 1 3 1 - 1 1 1 1 1 158 Cuna - - - - - 1 1 1 59 Goajiro 3 2 - 3 3 1 2 2 - 2 2 160 Haitians 3 1 2 3 - - 1 1 - 1 1 16 1 Callinago 2 2 1 3 2 - 1 2 - 1 2 162 Warrau 3 - - - 2 - 2 2 - - 3 163 Yanomamo 3 2 2 3 2 2 2 2 2 2 1 164 Carib 3 165 Saramacca 1 2 1 3 1 - 1 1 1 1 166 Munclurucu 3 ' 1 1 3 3 2 1 2 - 1 1 167 Cubeo 2 2 - 3 - - 2 2 2 2 3. 168 Cayapa 3 - - - - - 2 2 169 Jivaro 1 2 2 1 3 1 1 2 2 1 1 170 Amahuaca 1 2 - 3 3 - 2 2 1 71 I nca 2 1 1 1 2 1 1 1 - 1 1 172 Aymara 2 2 1 3 2 1 2 2 - 1 173 Siriano 2 2 - 3 2 2 2 2 - 2 3 174 Nambicuara - 2 2 3 - - 2 2 - 2 2 175 Truma i 2 2 - 3 2 2 2 2 2 3 3 176 Timbira 2 1 1 3 1 1 2 2 1 1 2 1 77 Tupinarnba 1 2 1 3 3 - 1 2 - 2 1 178 Botocudo 2 - - 3 3 - 2 2 - 2 179 Shavante 3 1 1 3 3 2 2 2 - 2 2 180 Aweikoma 2 2 2 3 3 2 2 2 1 3 3 181 Cayua 1 2 - 3 - - 2 2 182 Lengua - 2 1 3 1 1 2 2 2 2 1 183 Abipon 2 2 1 3 - - 2 2 2 2 1 184 Mapuche 1 2 1 3 3 1 1 2 - 1 1 185 Tehu.elche 2 2 1 3 2 - 2 2 - 2 2 186 Yahgan 3 2 2 2 1 - 2 2 2 3 479 Variable 34 35 36 37 38 39 40 41 42 43 44 -- -- 14 5 Creek 1 1 4 4 2 2 2 1 146 Natchez 1 1 4 4 2 2 1 2 1' 147 Comanche 2 1 1 1 4 4 2 2 1 1 1 148 Apache 1 1 1 1 1 2 2 1 1 2 149 Zuni 1 2 2 4 4 2 2 2 1 2 1 50 Havasupai 3 3 2 2 2 2 2 2 1 1 51 Papago 1 1 2 2 2 2 2 2 2 1 ·152 Huichol 3 2 3 2 2 1 53 Aztec 2 1 1 1 1 1 2 1 1 1 54 I>opoluca 2 1 1 1 55 Yucatec Maya 2 2 4 4 2 2 1 1 1 1 56 Miskito 1 2 1 2 1 2 1 57 Bribri 1 1 1 2 2 1 1 1 58 Cuna 2 2 1 2 2 2 2 1 59 Goajiro 2 2 2 2 2 2 2 1 2 160 Haitians 3 1 1 2 2 1 2 2 2 2 1 61 Callinago 2· 1 1 1 2 2 2 1 1 162 Warrau 2 2 2 2 2 2 2 2 2 163 Yanomamo 1 1 3 3 2 2 2 1 2 164 Carib 2 2 2 2 2 2 2 2 2 165 Saramacca 1 1 1 1 1 1 1 2 166 Mundu.ru.cu. 2 1 1 1 1 1 2 2 2 1 1 167 Cuoeo 2 2 3 3 3 2 2 2 1 2 168 Cavaoa 3 2 2 2 2 2 .L 2 ~ 2 169 Jivaro 1 1 1 2 2 2 1 1 170 Amahu.aca 4 4 2 2 2 2 2 1 71 Inca 2 1 2 2 1 1 1 1 2 1 1 172 Aymara 1 2 4 4 2 2 2 1 2 173 Siriono 3 2 3 4 4 2 2 2 1 2 174 Nambicu.ara 2 2 4 4 2 2 2 1 2 175 Tru.mai 3 2 3 4 4 2 2 2 2 2 1 76 Timbira 1 2 2 2 2 2 2 2 2 2 177 Tu.pinamba 1 1 1 1 1 2 2 2 1 1 1 78 Botocudo 4 4 2 2 2 2 2 179 Shavante 1 1 2 2 2 2 2 2 2 180 .Awe i koma 3 1 1 4 4 2 2 2 1 2 181 Cayu.a 2 3 4 4 2 2 2 2 2 182 Lengua 2 2 2 2 2 1 2 18 3 Ah i pon 3 2 ·1 1 2 2 2 1 2 184 Mapu.che 1 1 1 1 2 2 2 1 1 185 Tehuelche 2 2 2 2 1 2 186 Yahgan 2 2 2 2 2 2 2 480 Var:Lable Popul at ion 45 46 47 Density 145 Creek 2 2 3 146 Nat chez 1 2 4 147 Comanche 1 2 1 148 Apache 1 2 1 149 Zuni 1 1 3 1 50 Havasupa i 1 2 1 1 5 1 Papago 1 2 5 1 52 Huichol 3 1 53 Aztec 2 7 1 54 Popol uca 5 1 55 Yucatec Maya 1 1 2 1 56 Miskito 2 2 2 4 1 57 Bri bri 1 2 2 1 5.g Cuna 2 2 2 5 1 59 Goa jir o 1 2 2 2 160 Hai t i ans 2 2 2 6 1 61 Callinago 1 2 2 4 162 Warrau 1 2 2 2 163 Yanomamo 1 2 2 2 164 Carib 2 2 2 1 165 Saramacca 1 1 2 4 166 Mundurucu 2 2 2 1 167 Cubeo 1 1 2 2 168 Cayapa 2 2 1 3 169 Jivaro 1 2 1 2 1 70 Arnahuaca 2 2 1 1 71 I nca 2 1 2 4 172 Aymara 1 1 1 6 173 Siriono 1 1 2 1 174 Nambi cuara 1 2 2 2 1 7:5. Trumai 1 1 2 1 1 76 Timbira 1 2 3 17 7 Tupinamba 1 2 2 2 1 78 Botocudo 1 2 2 1 179 Shavante 1 1 2 1 180 Aweikoma 1 2 1 1 1 8 1 Cayua 2 1 2 2 182 Lengua 1 1 2 1 183 Abipon 1 2 2 2 184 Mapuche 1 1 2 5 185 Tehuelche 1 2 2 1 186 Yahgan 1 2 2 2 481 APPENDIX D Comparison of Phi Coefficients Among Geographical Regions Vvorld Africa C-Med Eurasia Pacific N Amer S Amer Variable 2 with variables: 2 l .159 .2 25 .500 -.07 8 -.2 42 .159 .333 22 .225 . 55 6 .207 .096 - .171 .270 .223 23 .081 .256 .034 -.254 -.355 .294 .068 24 .026 .507 - .187 -.234 -.088 .139 -.259 25 - .111 - .181 - . 2 19 . 410 .089 -.053 -'.224 26 .039 -.081 . 140 - .140 .1 23 -.017 - . 01 7 27 - .144 . 01 8 -.283 -.346 - .134 -.028 -.225 28 -.084 -.333 . 316 0 0 - .12 1 -.238 29 .11 2 .408 .1 68 - . 05 0 - .123 .269 • 01 8 30 . 014 .242 .1 68 - .141 -.071 .237 -.208 31 -.030 .250 - .1 34 -.461 -.049 . 45 8 . 214 32 -.228 - .132 .257 -.093 -.402 -.290 -.256 33 .26 8 .429 .2 40 .233 .197 .093 -.032 34 -.094 0 -.444 0 .350 .333 . 791 35 . 25 1 .478 .255 -.207 .237 .279 0 36 .27 6 .522 .122 - . 01 0 .252 . 214 . 616 37 . 321 . 60 1 .282 .406 -.078 .1 77 .537 38 .176 .45 4 .067 .164 -.023 -.078 .114 39 -.011 .1 97 -.0 67 . 311 .023 -.425 - . 015 40 -. 110 .234 - .1 82 - .187 -.282 0 - .1 04 41 .056 .23 4 . 07 1 -.078 - .168 - .137 . 01 4 42 .175 .207 .355 - . 013 .193 -.237 .1 0 4 43 .30 4 .428 .588 .204 .049 -.099 .244 44 .237 .399 .5 44 .20 8 .039 .1 68 .078 45 .097 0 .240 -.050 .023 0 .389 46 -.043 -.099 -.342 -.330 .205 . 01 6 .035 47 -.023 . 11 0 - . 2 7 1 0 .044 -.207 .1 53 Variable 3 with variables: 2 1 . . 069 -.040 0 - .1 24 -.226 .1 07 .492 22 . 0 1 9 .604 0 .1 91 - .171 - .309 - .1 31 23 - .042 .404 0 -.045 -.509 - .183 .096 24 - . 010 . 11 7 0 -.060 - .1 67 - . 03 1 - .1 82 25 - .183 - .1 77 0 0 - .1 06 - . 415 -.523 26 .1 79 - .1 65 0 .259 ·- . 036 .309 . 2 7 1 27 -.009 -.306 0 .122 -.094 -.033 - . 2 0 1 28 - .007 - .125 0 - • 2 1 8 0 - .118 .100 29 . 0 1 8 .240 0 - .1 66 - .145 .145 -.220 48 2 30 .070 .354 0 -.202 - . 218 .043 -.042 31 -.026 .234 0 .043 .107 - .169 -.730 32 - .123 -.362 0 .037 -.424 -.291 -.045 33 .129 .165 0 .064 .039 .119 .15 6 34 .103 .239 0 .632 .300 -.300 .15 8 35 -.035 0 0 - .189 -.222 -.074 -. 228 36 .062 .182 0 - .164 .123 .267 - .126 37 .115 . 215 0 .347 0 -.017 -.095 38 -.009 .025 0 .227 -.073 -.007 -.292 39 -.064 - . 061 0 .019 -.258 . 316 - . 421 40 .095 .167 0 .17 6 - . 312 0 -.033 41 -.078 .198 0 -.445 -.470 .163 - .149 42 .lll .289 0 - .13 7 .159 .043 -.022 43 .010 .222 0 .092 • 015 - . 081 -.098 44 .045 .040 0 .007 . 065 .320 -.042 45 - .108 -.079 0 -.246 .275 0 -.040 46 .024 0 0 - • 091 .260 -.208 .083 47 -.075 0 0 0 -.035 -.337 .267 Variable 4 with variables: 21 . ll5 . 331 .327 .078 - .121 .13 8 .076 22 .029 . • 415 - .127 .115 · . 015 .189 - .155 23 -.027 .205 - . 221 - .120 -.266 .299 .042 24 .036 .060 -.302 .067 • 217 .245 . 219 25 -.018 .200 -.256 .225 - . 218 .020 .230 26 .024 .243 .184 .096 -.020 .096 -.184 27 -.011 .362 0 -.363 .247 -.273 - .109 28 .037 .556 .189 -.258 0 -.233 .033 29 0 .089 -.064 .141 0 .068 -.080 30 - .108 -.053 -.064 .033 - .150 -.375 - • 015 31 -.070 .19 6 -.356 .243 - .141 .354 - • 214 32 -.097 - .132 -.385 .090 -.066 .109 -.017 33 .008 .171 .359 0 - .187 .066 - . 210 34 -.030 .284 -.577 0 .598 . 218 0 35 .1 7 l .679 .444 .14 6 -.237 .15 7 .189 36 - .088 .380 -.353 -.112 - .182 .033 - .122 37 - .116 .060 - .190 - .164 -.250 .132 - .102 38 - ·.13 8 -.029 -.258 .10 4 -.082 .163 - . 253 39 -.058 -.007 -.086 -.048 -.183 -.075 - . 067 40 - .16 7 -.062 -.232 .078 -.275 - 0 -.046 41 -.004 .20 8 -.229 .230 .124 . 051 .144 42 .00 4 -.055 .063 -.204 .044 -.375 .ll5 43 -.009 .257 .071 .ll4 - . 013 .164 - .142 44 .034 .150 .247 .050 -.053 .229 - .142 45 .171 .29 3 .359 .12 2 .135 0 -.058 46 .143 .372 - .199 0 .173 . ll2 .349 47 -.241 - . 410 -.598 - . 019 .155 .077 -.354 483 Variable 5 with variable s: 21 -.036 . 071 -.354 .433 .200 -.545 -.356 22 .046 . 241 - .172 .149 .182 0 .117 23 .049 .123 - . 2 63 .107 .30 9 .299 - .139 24 .03 4 .23 8 .292 .0 86 ·• 016 -.234 .259 25 .09 5 .139 .464 .015 -.047 . 212 .154 26 .022 - .174 .472 . 015 -.026 .162 - .140 27 .027 - • 05 6 .224 .25 9 -.141 .113 -.046 28 -.049 0 -.447 .293 0 -.073 -.267 29 . 061 .19 6 .146 - . 021 .102 .127 .168 30 .003 - .116 .146 .046 .101 .191 .237 31 .092 . 612 -.478 .386 -.039 - . 411 .356 32 .147 .183 .408 .089 .284 .204 .181 33 .040 .337 -.329 .247 .007 . 011 • 013 34 -.303 .184 -.350 • 316 -.350 -.655 -.500 35 .083 .329 -.270 .258 -.237 . 067 .125 36 .160 .704 - .181 .243 . 0 61 .042 -.099 37 .089 . 489 -.232 .005 .060 • 3 91 -.069 38 .053 - . 061 .090 .202 .204 -.055 .115 39 .076 - . 012 - .161 .059 .289 -.022 .106 40 .0 86 .181 .343 -.004 .233 0 .185 41 .069 .181 .145 -.030 .139 .097 .138 42 -.033 - .13 7 -:-. 071 .. 160 -.110 .191 .111 43 .072 . 331 - .194 .036 -.083 .110 .200 44 .093 .187 .187 -.043 - .167 · • 089 .050 45 -.038 - .154 - .161 - . 219 - .113 0 -.248 46 -.060 - . 012 • 241 - . l?. 6 -.083 - .159 -.022 47 .022 - . 216 .192 0 - .106 -.233 .163 Variable 6 with variables: 21 - .12 8 -.042 .149 . 206 -.256 -.328 .0 66 22 -.047 - .111 .142 .171 -.308 -.270 .036 23 - • 081 - .193 .239 -.050 -.3 89 -.240 -.084 24 -.288 - .141 -.688 -.476 - .162 -.233 - .123 25 -.274 I - . 014 - .156 - .19 6 -.221 -.334 - .155 26 - .102 . 081 -.022 -.275 .32 2 -.009 -.042 27 -.044 - .16 7 .050 .232 - • 312 - . 210 .191 28 - • 061 .17 4 .378 . 408 0 - .191 - . 071 29 -.209 0 .076 - .188 -.408 -.064 - .632 30 - .1 91 -.040 .076 - .145 - . 2 91 - .141 -.429 31 -· .129 .075 0 - • 051 - . 2 62 -.258 .069 32 -.420 -.23 6 -.535 - . 216 -.247 -.380 -.637 33 - .05 6 -.079 .35 8 - .153 .187 . 091 -.565 484 34 - . 033 . 640 - .0 59 -1. 00 -.395 . 447 0 35 - .127 -.265 .10 0 - .141 .105 .095 - .4 72 36 - .133 .058 - . 313 -. 259 .015 -.040 - .17 9 37 -.ll5 - .17 6 -.053 - .128 - .1 81 .0 84 - .095 38 - .181 -.279 -.060 .068 - .13 6 -.267 -.368 39 -.099 -.369 -.071 -.063 .13 6 .0 80 - .12 0 40 -.278 . 019 -.482 -.272 - .155 0 -.333 41 - . 081 . 019 . 079 . 094 · -.092 -. 282 -.249 42 . 015 - . 015 . 482 .015 -.192 -.141 -.200 43 .094 - .196 .3 02 . 410 - .174 -.030 -.067 44 -.064 - .207 . 462 -.253 -.246 -.llO -. 447 45 . 208 - .3 65 . 358 .188 -.050 0 .447 46 - • 013 .099 .02 9 -.230 .055 .033 -.067 47 - . 061 - .llO - • 219 0 • 25 6 - .106 -.09 8 Variable 7 with variable s: 21 . 111 . 25 6 -.ll5 -.297 - . 018 -.066 .204 22 -.084 0 -.266 -.383 . 061 - .180 -.025 23 . 019 .30 8 - .14 7 . 01 2 .135 . - . 239 - .141 24 - . 151 -.256 .134 - . 291 -.088 - .177 -.332 25 -.034 - . 316 - .15 6 -.225 • 2 81 0 ...: .141 26 .070 0 - . 131 -.096 .1 23 0 .067 27 .127 .596 .050 .199 .116 0 -.059 28 .129 .293 0 0 0 0 .141 29 - .11 1 - .167 -.0 62 -.251 -~20 8 0 .13 7 30 - .178 -.230 -.0 62 -.227 -.279 0 - .142 31 -.284 - .0 68 -.389 0 - • 519 0 - .179 32 -.043 - . 378 -.268 -.093 -. 065 0 .225 33 - . 018 -.357 .0 66 - .190 - .107 - .125 .172 34 -.003 -. 083 -.059 0 -.0 60 0 .500 35 .07 0 .094 - .0 60 - .207 - .139 - • 212 .250 36 -.056 -.032 .024 - .310 .023 - . 216 .058 37 -.083 - . 280 -. 205 -.268 .00 8 - .155 .13 2 38 -.027 - . 271 .147 - .152 -.268 -.077 .167 39 - .007 - .145 . 104 -.365 - .1 64 .130 .253 40 -.089 - . 118 .1 81 - . 219 -.306 0 -.078 41 -.078 - .118 -.234 - . 091 .112 0 - • 201 42 - .17 4 -.077 -.395 -.079 -.204 0 - .1 62 43 -.089 -.015 -.320 -.270 - .19 3 0 .054 44 ' . 011 .089 - .182 - .1 33 - . 016 0 . 281 45 .088 . 474 .0 66 - .120 . 230 0 .040 46 - .184 .037 - .178 -.204 -.344 0 - .197 47 - .036 - .158 - .154" 0 -.228 0 .184 485 Variable 8 with variable s: 21 .03 4 -0220 -.044 - . 017 .177 .• 3 61 -.238 22 .111 .098 .068 .089 .171 .348 - . 014 23 .062 - .13 0 - .107 .038 .145 .347 . 221 24 .392 .378 .500 . 62 9 .184 .283 .411 25 .277 .309 .250 .309 - .129 .334 .279 26 . 041 -.058 .085 .308 -.302 .009 -.017 27 -.060 -.472 -.082 - . 311 • 061 . 210 - .145 28 -.055 -.408 - • 316 -.408 0 .191 - • 051 29 .286 .167 -.023 .340 .444 .064 • 518 30 .323 . 2 63 -.023 .284 .439 .141 .567 31 • 351 0 .389 • 051 • 651 .258 .134 32 .431 .548 .675 • 25 6 .197 .380 .462 33 .069 .408 -.370 .280 - • 017 -.047 .444 34 .030 -.386 • 091 1 • 00 .316 -.447 -.500 35 .062 .15 2 -.025 .244 .080 -.015 .15 8 36 .173 .082 . 2 61 .459 -.032 .141 .123 37 .17 5 • 415 • 201 .305 .115 -.027 -.024 38 .194 .524 -.053 .034 • 336 -.305 .209 39 .100 .480 - • 012 .296 .070 - .133 - .120 40 .335 .1 0 l .300 .399 .383 0 .413 41 .140 .101 .099 -.030 -.048 .282 .444 42 .127 .089 - .144 .036 . 310 .141 .357 43 - • 016 .185 -.037 - . 211 .289 .030 • 017 44 .052 .089 -.282 .324 .168 .11 () .190 45 -.268 - .15 8 -.370 -.099 - .185 0 -.494 46 .161 - .123 .1 04 .349 .289 -.033 .255 47 .087 .253 .309 0 .055 .106 - •. 075 Varia ble 9 with variables: 21 -.085 -.118 0 -.35 4 • 013 .200 -.356 22 - .139 - .13 8 -.068 -.050 -.105 - . 311 .149 23 - .116 -.018 .107 - • 314 -.023 -.366 .053 24 -.34 8 0 -.40 8 -.385 - • 211 -.459 -.222 25 -.188 . 061 -.259 .055 - . 218 -.478 -.038 26 -.075 - .12 2 - • 212 • 015 .112 - .189 .141 27 - .12 7 -.342 -.20 4 .160 -.0 60 - • 365 .194 28 -.00 8 -.429 .500 .25 8 0 - .149 -.267 29 -.239 - .150 - . 216 -.231 -.068 -.208 - . 281 30 -.255 -.03 2 - . 216 - . 265 .101 . -.337 - . 3 21 31 -.208 -.055 .2 89 -.098 -.408 -.033 0 32 -.268 .123 - . 4 71 -.359 - .17 5 -.284 -.247 33 -.004 .20 6 .055 .088 . 081 -.094 - . 3 71 486 34 .077 .0 83 . 055 0 - • 316 .333 .250 35 - . 132 -.09 4 .158 -.337 - .15 4 - . 437 -.094 36 - . 212 - • 391 - .15 8 -.287 - .105 .13 6 - .3 80 37 -.037 - . 094 .050 • 081 .181 -.010 -.009 38 - . 117 - . 140 - . 16 8 - .12 7 - . 043 - . 183 .1 63 39 . 002 - . 263 . 012 -.200 . 289 . 122 .046 40 - • 312 - . 158 - . 378 - . 244 - . 218 0 -.234 41 - . 138 - . 158 - .123 - .133 . 139 -.369 .155 42 - .1 29 0 - . 098 - .115 . 288 -.337 - .207 43 . 085 -.079 .13 7 .16 4 . 089 .1 0 1 .124 44 .037 - .1 83 .10 1 . 219 • 012 -.064 -.092 45 .1 0 1 - . 354 . 0 61 -.040 .264 0 .277 46 - .149 -.060 • 013 - . 165 -.256 -.330 • 041 47 - . 151 .14 7 -.28 6 0 - .106 - .189 -.303 Variable 10 with variables: 21 - .03 4 - . 411 0 - .131 . 221 -.07 5 .131 22 - . 195 - . 417 -.187 - .149 . 435 -.226 - . 411 23 - .13 8 -.5ll - .0 67 - .17 8 . 241 - • 301 -.025 24 -.238 - .12 6 - .355 -.333 .1 62 -.062 -.520 25 - . 254 .132 -.233 - .19 6 - .13 4 -. 240 -.533 26 - .140 - .130 -.040 -. 275 -.172 -.200 - .141 27 - .143 0 - .3 54 . 354 - .042 -. 474 - .0 60 28 .040 0 .250 .48 8 0 - .026 -.071 29 - . 280 - . 423 -.062 -.257 .129 -.533 -.322 30 - . 234 - . 295 - .0 62 -. 335 .154 -.244 -.335 31 -.069 - . 145 0 -. 051 .0 94 -.141 -.069 32 -.292 • 254 - • 514 -.470 -.344 - • 091 -.324 33 - .1 2 8 - .33 6 .175 - . 153 .2 82 -.247 - .3 89 34 -.0 60 .433 .068 -1 . 00 -.598 -.218 .189 35 - • 219 - . 478 - . 040 -.141 .19 6 -.238 -. 403 36 - .13 0 - .411 -.224 - . 310 .367 .18 7 - • 414 37 - .15 0 -.208 - . 178 -.289 . 321 -.084 -.233 38 - . 148 - .35 6 .0 29 - .13 3 .008 - .115 -.254 39 -.045 -.273 .168 - .123 - . 00 8 . 329 -.376 40 - . 244 - . 169 - . 461 - . 30 4 - . 044 0 -.232 41 - .02 4 - . 169 .2 22 .077 . 482 - .143 -.302 42 - .175 - . 293 . 247 -. 288 .058 -.244 -.243 43 . 020 -.507 .118 . 227 .14 7 . 258 - .17 8 44 - .14 7 - . 293 .114 -.079 .099 - • 621 - . 108 45 . 084 -.378 . 4 61 .007 - .20 2 0 .253 46 - .0 56 .064 . 284 -.290 • 310 - .189 - • 081 47 .05 5 - .079 - . 154 0 .33 8 . 076 .0 24 487 Variable 11 with variable s: 21 - .047 .1 61 0 - .189 - .181 .2 38 -.408 22 .059 .15 8 .09 G . 103 - . 471 -.029 .550 23 .025 • 371 .189 - .1 07 - . 236 . 012 .068 24 - .10 4 .084 -.164 -.027 - . 314 -.323 . 35 8 25 .040 0 -.10 9 .220 -.040 -.169 .452 26 . 058 -.060 -.208 .279 .233 .032 -.042 27 .00 9 - .342 .070 - .169 - . 012 .15 8 .191 28 -.042 -.429 .378 -.293 0 - .109 -.071 29 .03 5 .096 -.196 .043 -.165 .344 .103 30 - • 021 .13 7 - .196 .090 -.050 .043 .083 31 - .133 .026 .2 89 -.051 -.420 .098 .069 32 .0ll • 010 - .102 .077 .168 - .14 7 .144 33 .ll5 .436 -.099 .23 6 - .169 .15 7 .13 7 34 . ll3 - .15 8 0 1 . 00 • 316 .447 0 35 .072 .267 .209 - .149 -.277 - .12 6 .322 36 -.070 -.045 .03 5 . 031 -.385 -. 071 .154 37 .105 .060 .258 • 344 -.164 .073 .244 38 .029 .131 -.225 • 015 -.042 -.045 .410 39 .044 -.045 - .143 - • 061 .243 -.206 • 441 40 -.0 63 -.037 -.042 .077 - .139 0 • 051 41 - .10 5 -.03 7 -.332 - .197 -.302 - .167 .444 42 .043 .19 4 -.324 .17 9 .184 -.043 .085 43 .0 60 .262 .05 8 -.074 -.053 - .167 .28 9 44 .1 7 1 .022 • 019 .276 -.075 .549 .038 45 .016 -.084 -.324 -.043 .3 85 0 -.038 46 -.0 86 -.099 -.228 .13 6 -.476 - . 091 .119 47 - .192 .19 2 - .198 0 -.377 -.235 -.274 Variable 12 with variables: 21 - .ll4 - .134 -. 462 .307 -.375 - . 019 -.036 22 - . 15 9 -.053 -.233 -.266 - .105 -.289 - .120 23 -.079 - .13 0 -.250 -.074 .0 23 -.293 .0 68 24 .049 .134 . 402 -.030 .076 .040 - • 217 25 -.055 .106 • 210 -.062 .185 - • 212 -.300 26 - . 031 -.289 .0 94 - . 210 .14 6 -.200 .503 27 -.050 -.286 .141 . 363 - .149 -.304 .178 28 - . 010 -.40 8 -.333 .333 0 .233 .141 29 -. 035 .268 .0ll • 010 -.080 -.223 -.071 30 - . 013 .44 5 . 0ll .034 - .145 -.240 -.005 31 -.076 -.250 -.375 .141 .094 -.240 -.069 32 - .0 05 .108 .245 - .138 .178 -.226 -.040 33 .026 0 -.308 .076 .079 0 .17 2 488 34 - .159 .030 -.033 0 -.395 _ -.218 -.250 35 - . 062 - .189 - .149 .372 0 .005 -.250 36 -.165 -.522 -.344 - . 010 .015 .040 -.285 37 - .144 -.078 -.407 -.268 -.060 ,- .053 - .143 38 .065 .006 .042 .062 .130 -.078 .114 39 . 051 - .12 2 .129 .249 .160 . 014 - . 015 40 .160 .222 .352 .316 .082 0 . 081 41 -.043 - . 018 -.059 - .104 -.109 - .107 . 014 42 -.078 .470 - .146 - .133 -.226 -.240 - .174 43 -.045 - .144 -.508 . 041 -.070 .13 0 -.035 44 -.102 -.259 -.229 - .152 -.080 - .181 .078 45 - . 051 -.087 .103 -.010 -.429 0 -.078 46 -.067 -.007 .145 - .108 -.070 - .144 - .104 47 . 301 .330 .588 0 . 5 21 .193 .153 Variable 13 with variables: 21 . 001 -.299 .008 . 2 61 - . 013 -.094 -.357 22 .231 .204 .110 .286 - . 015 .258 .203 23 .039 - .13 0 - .15 7 . 291 -.220 .223 .207 24 .467 . 462 . 5 64 . 651 .438 .109 .144 25 .232 .032 .490 .120 .047 .258 .218 26 .159 - .182 -.039 . 464 .020 .352 .350 27 .059 -.263 .394 - .100 .060 .273 . 211 28 -.273 - . 816 -.333 -.683 0 - .136 .302 29 .258 .272 -.103 .269 .238 . 213 .327 30 .3 47 .342 - .103 • 316 .235 .089 .484 31 .340 .354 - .167 .071 . • 535 0 .418 32 .533 .462 .58 8 . 614 .455 .680 .333 33 . 061 .447 - .177 .162 -.262 .359 .204 34 -.025 - .184 . 051 .250 -.060 - • 218 -.250 35 • 0 66 .033 - .149 .258 .196 .234 .185 36 .224 .066 .589 .422 -. 061 .267 -.034 37 .252 .. 272 .468 .251 .037 .309 - .182 38 .296 .597 .336 .17 8 .043 .244 .378 39 . 031 .267 - .103 .162 -.043 -.085 .161 40 .560 .586 .590 .342 .668 0 .599 41 .373 .58 6 .15 9 .055 .197 .694 .802 42 .109 .15 0 -.406 .164 -.288 .089 .466 43 -.039 .355 -.259 -.352 .083 .154 .209 44 -.051 .320 - • 451 . 081 - . 012 .229 .145 45 - . 2 61 - . 081 -.275 · - . 015 - • 264 0 - .145 46 .069 - .195 -.454 .410 .083 - .112 .354 47 .045 .275 .193 0 .106 -.077 -.086 489 Variable 14 with variable s: 21 .108 .30 8 . 2 71 .175 -.027 -.043 - .191 22 .276 .47 6 .5 86 . 21 4 - .17 9 .405 .208 23 .184 .1 93 .3 93 .116 - .138 .50 9 . 348 24 . 305 .38 8 .243 . 412 .027 .139 • 35 6 25 .27 4 . 091 0 .187 .204 .246 .37 8 26 .182 - . 081 0 .30 8 .220 -.017 .135 27 - . 014 - .12 2 0 -.277 . 071 - .105 -.022 28 -.043 -.333 0 -.488 0 -.185 .302 29 . 436 .497 .406 .351 .247 .418 . 612 30 .407 .30 9 . 406 .303 .2 22 .084 .709 31 .2 46 .302 0 .2 25 . 430 -. 228 . 418 32 . 450 .13 2 0 • 646 .576 . 5 51 .363 33 .209 .553 .302 -.042 -.257 .513 .439 34 .175 . 284 0 • 612 • 316 .149 -.378 35 .259 .564 .333 - .189 .0 26 . 476 .410 36 . 214 .592 . 513 .135 .032 • 214 . 201 37 .3 23 .478 . 519 .09 7 • 313 .465 - . 061 38 .17 4 .330 • 221 .12 0 .159 - .104 .354 39 -.009 .12 2 -.202 • 214 - .159 -.080 . 113 40 . 353 .259 .192 .289 .411 0 • 451 41 . 228 .259 .120 -.090 -.0 28 .2 60 .668 42 . 290 . 259 .2 27 .2 67 .0 22 .0 84 .536 43 .10 3 .5 84 . 295 -.267 .2 28 . .070 .222 44 .127 .442 .0 93 .263 .070 • 315 . 2 71 45 -.28 8 .087 -.20 9 - . 351 -.223 0 - • 271 46 . .151 - .169 - .147 .343 .0 87 .173 .149 47 - .101 .121 • 063 0 - .149 - .191 - .160 Variable 15 with variables : 21 - .118 - .149 0 .048 -.278 • 281 0 22 - .15 6 -.3 64 0 .058 -.3 67 - .329 .174 23 - .13 7 -. 205 0 - .148 -.389 -.294 .0 68 24 -. 298 -.3 34 0 -.437 - . 217 -.324 - .1 23 25 -. 26 0 .12 5 0 - .129 -.055 -.395 - . 091 26 - . 117 - .132 0 -.243 .250 • 017 -.042 27 - .151 -.2 48 0 .194 - . 461 -. 279 - . 017 ) 28 .016 .17 4 0 . 408 0 .2 63 - . 386 29 - . 2 61 -.060 0 - .165 - .183 - . 418 - . 360 30 -.234 .036 0 - .134 - . 211 -.084 - . 2 71 31 - .140 - .19 6 0 .141 - .12 7 -.258 - .17 9 32 -.3 95 - . 218 0 -. 465 - .149 - . 341 -. 480 33 - .153 - . 289 0 - . 01 6 -. 007 - . 134 -.389 490 34 -.239 0 0 -1.00 -.060 -.57 7 0 35 - . 271 -. 478 0 -. 026 -.294 -. 386 -.351 36 - .047 -.380 0 - . 010 - • 061 .13 7 - . 066 37 - .166 -.442 0 - . 109 - .2 92 .004 • 051 38 - .130 -. 320 0 -.006 - .1 83 - .02 2 -.07 9 .39 -.077 - • 291 0 - . 218 -.082 .119 - .120 40 -.305 - .15 4 0 -.30 6 - .20 8 0 - . 333 41 - .122 - .154, 0 .182 - .124 -.260 -.028 42 - .1 06 -.010 0 -.07 2 -.044 -.084 - . 200 43 .150 -. 435 0 .537 • 013 • 2 69 . 067 44 - . 071 -.263 0 - .18 6 -.139 - • 315 -.298 45 . 282 - .12 0 0 . 313 .067 0 .298 46 -.111 . 066 0 -.103 .013 -.330 -.200 47 - . 016 -.072 0 0 .145 -.08 3 -.098 Variable 16 with variables: 21 . 054 .094 .101 - . 201 -.468 - .194 .524 22 . 077 . 65 8 - .196 .15 8 .149 - . 810 .076 23 -.0 68 . 316 - .182 - .115 - . 333 -.756 -.029 24 .0 74 .362 -.098 .077 -.059 -.463 .179 25 - .100 - .139 -. 234 .050 - .19 2 - • 218 -.272 26 -.005 - . 451 .25 8 . 258 -.507 .667 0 27 -.045 -.23 6 - .134 .1 60 • 381 .327 - . 671 28 - .22 6 - .19 4 -.200 -.354 0 .218 - . 548 29 . 065 . 402 .294 -.037 -.016 - .102 - .169 30 .111 .3 25 .29 4 -.083 - • 221 .218 -.293 31 - .140 .13 5 0 -.060 .577 -. 44 7 -1.00 32 - .130 -.299 -.043 - .03 3 -.447 -.327 - . 239 33 .193 .3 87 . 13 9 .177 - .143 .083 . 408 34 . 091 0 - . 378 0 -1.00 0 0548 35 .0 85 0 • 655 .37 8 0 - • 218 -.250 36 . 081 .333 - . 258 -.055 0 .1 89 . 333 37 .1 85 .443 - . 213 .3 27 -.0 69 -.327 . 478 38 .0 85 .262 - • 218 .204 .22 8 . 241 - • 314 39 -.096 .055 - . 293 .0 83 -.424 . 2 61 -.378 40 .0 57 .149 . 237 .27 8 - . 419 0 - . 529 41 - .143 .118 - . 207 -. 405 -.419 0 - .357 42 .181 .2 93 . 207 -.2 25 .367 • 218 - . 07 6 43 . 021 .322 - .154 -.0 58 .071 .1 67 .076 44 .0 62 .233 . 237 - . 03 6 • 491 0 .029 45 - • 241 -.262 .080 -.294 .073 0 -.255 46 .0 24 -. 262 - . 237 .00 9 . 464 -.375 . 314 47 - .047 - . 081 0 0 . 250 - .2 50 .255 49 1 Variable 17 with variable s: 21 .26 8 .30 8 . 458 .267 .279 -.0 66 .16 7 22 .12 4 .05 3 .0 85 .0 86 - .17 9 .197 . 313 23 .117 . 476 . 331 .044 .0 55 .0 84 - .127 24 .011 • 214 -. 625 -.033 -.093 .093 .182 25 .0 90 • 014 -.340 -.076 - .120 .246 .392 26 -.017 - . 081 - . 214 -.031 .220 - • 017 - .300 27 . 0 61 . 018 - 0138 - .149 0 .279 .146 28 .02 8 - .17 4 .500 -. 488 0 .121 • 071 29 .008 -.07 6 -.043 -.0 68 -.082 - . 031 .184 30 -.006 .121 -.043 - .10 8 - . 2 61 -.0 68 .046 31 -.053 -.167 .35 6 -.283 -.209 .125 .449 32 .112 -.325 -.3 28 • 216 - .144 . 341 • 381 33 .099 .171 .500 -.278 .196 - . 091 -.038 34 .15 7 - .178 .167 .632 .100 .333 .378 35 . 291 .478 .350 -.0 67 . 419 .190 . 4 71 36 - .106 -.050 -. 071 - . 216 -.105 -.29 9 .328 37 .028 .17 6 .165 -.033 -.038 - .16 7 .087 38 - . 041 -.006 -.25 8 -.068 -.243 • 051 .160 39 -.088 -.043 -.043 -.063 - .159 -.307 . 216 40 -.10 2 . 01 8 -.564 - .155 - .107 0 .022 41 -. 003 -.222 • 019 -.0 94 .263 - .13 7 -.028 42 . 169 . 259 .5 64 .11 6 - . 2 61 -.0 68 .33 3 43 .119 .144 .549 - . 015 -.251 .240 .333 44 .059 - .105 .373 .253 .120 - .12 7 .298 45 .121 . 434 .350 .323 .040 0 .29 8 46 . l--16 .182 .073 .230 - .111 .173 .333 47 -.0 85 .121 -. 395 0 - .125 . 3 72· -.098 Variable 18 with variables : 21 - .134 .040 -.3 93 -.0 27 -.040 .263 .02 4 22 .002 -. 407 . 016 .13 9 .0 26 -.0 66 .110 23 . .0 84 .15 9 0 .0 69 .0 30 -.084 .330 24 -.064 - .280 .140 - ,19 4 -.053 -.0 93 - .121 25 - .0 88 -.298 . 250 - . 204 .27 3 -.246 .05 8 26 - • 01 9 .2 78 .13 9 - • 411 .13 9 .0 42 -.053 27 -.088 - .2 86 .250 0 - .146 - .12 5 - .207 28 -.28 9 - . 241 - ·. 632 -. 408 0 -.284 -.071 29 .0 92 - .0 26 .260 -.008 .1 85 -.02 5 . 0 65 30 . 081 .14 6 .2 60 .045 .045 - .180 .0 89 31 .072 -.250 . 102 .2 39 0 . 043 - .17 9 32 - .0 54 -.220 .1 09 -.007 .0 98 -.099 -. 066 492 33 -.078 .086 - .13 7 -.078 -.090 -.045 .145 34 -.285 -.365 -.458 -.632 0 -.333 0 35 - .1 69 -.357 . 2 71 .2 58 - .2 84 -.3 33 - .167 36 .029 -.290 -.094 .195 - . 3 21 .2 99 .025 37 -.089 - .17 6 - .118 -.027 -.226 - .11 3 .113 38 -.007 -.058 - .1 67 .150 -.042 .130 • 031 39 .033 .099 -.380 .149 -.029 .446 -.090 40 .140 .055 . 404 .025 .108 0 .080 41 -.055 -.218 - .177 .20 6 - .184 .148 -.045 42 . 061 .250 . 081 .043 -. 014 - .180 .055 43 . 021 0 - .243 - . 012 - .122 .223 • 289 44 -.057 -.375 - .138 - .131 -.115 .0 65 .123 45 .084 .2 80 -.299 . 312 .050 0 -.032 46 - ·. 005 0 .040 - .149 -.236 -.064 .167 47 -.055 - .102 • 210 0 -.229 - • 061 - .120 Variable 19 with variables: 21 .078 - .144 0 .378 -.053 .306 -.192 22 .228 . 338 .164 .366 - .171 .320 .236 23 .045 -.260 - .107 .220 -.055 .375 .238 24 .430 .300 .578 .578 .187 .373 .369 25 • 364 . 586 . 351 .266 .434 .13 7 .24 3 26 .227 -.058 .13 9 .378 .356 .101 .250 27 - • 015 -.258 .3 35 -.181 - .175 -.020 - .155 28 -.162 -.556 -.632 -.500 0 -.026 .267 29 .29 2 .230 . 216 .044 .368 .399 .208 30 .355 .351 • 216 .098 .296 .346 .389 31 .246 .452 -.356 - .100 .403 . 411 -.239 32 .502 . 818 .459 .552 .667 .043 .169 33 .11 9 . 408 -.144 .39 9 - .103 0 .397 34 - .108 -.386 -.043 .447 -.258 0 -.378 35 .086 . 426 -.055 .537 - .194 .101 -.040 36 .254 . 390 .478 .373 . 284 -.033 - • 018 37 .288 .344 .275 .387 .344 .246 -.300 38 .205 .457 .168 -.032 .090 .289 .23 6 39 .011 • 381 - . 2 67 .15 3 .042 -.232 -.024 40 .430 .282 .63 2 . 208 • 5 80 0 .199 41 .316 .498 .480 - .148 . 061 .370 .262 42 .149 .195 - .158 . 031 - .1 2 7 .346 .358 43 .059 .220 -.077 -.025 .106 .089 .243 44 -.004 .195 -.539 -.032 -.084 .429 .306 45 - .2 47 -.078 - . 220 - .172 - . 313 0 -.084 46 .110 .10 3 - .184 .240 -.042 .15 9 .18 8 47 .024 .234 .286 0 -.042 - .145 - .12 6 493 Variable 2 0 with variables: 21 . 014 - . 415 . 292 .1 2 5 .10 4 .0 57 - . 314 22 .100 .149 0 . 13 8 - . 321 .37 8 0 23 . 095 -.038 .3 46 .00 5 - . 314 .309 . 277 24 - .140 -.220 .156 - .19 6 - .175 - . 166 - .183 25 . 324 . 25 2 . .222 -.020 .262 . 436 .7 64 26 . 184 .290 0 -.306 .636 .372 -.340 27 0 -.478 .167 .414 -.333 .187 .149 28 .070 .293 . 612 .192 0 - .191 .250 29 - . 004 - . 102 - . 231 .053 • 016 .133 0 30 -.075 .1 61 - . 231 - .0 24 - . 13 8 - • 216 - .126 31 .101 . 408 0 . 069 .025 .1 67 -.060 32 .208 . 462 .238 -.044 .29 9 .444 -.066 33 .0 67 .13 5 - .15 6 - .151 .0 66 .173 .12 5 34 - . 019 .051 0 • 316 -.350 .577 - . 417 35 .076 .3 29 .076 - .125 -.059 .086 - .169 36 -.04 6 -.17 .9 .073 - .2 03 .03 6 .0 83 -.0 98 37 .103 - .164 . 311 . 068 .267 . 205 -.224 38 . 049 • 012 -.115 .042 - .25 6 .141 . 330 39 - . 018 -.074 -.204 .177 -.316 -.025 .145 40 .040 .118 - .045 -.005 .235 0 - .10 8 41 .146 .337 .32 8 -.245 - . 032 • 210 . 212 42 - . 041 .077 .15 6 -.005 . - . 321 - . 216 . 212 43 .202 .015 .489 .145 .0 86 .022 .458 44 . 072 .:. . 25 6 . 122 . • 377 0 .194 .120 45 - . 071 - .15 8 -.045 .07 5 -.272 0 .0 80 46 . 043 .443 - .191 -.0 24 - .135 -.295 . 4 71 47 - . 060 -.253 .22 4 0 - .158 .0 20 - .10 8 Variable 21 with variables : 22 .171 .048 . 207 .267 .171 .189 .140 23 .13 2 .182 .272 .2 39 . 270 - • 061 -.0 60 . 24 -.083 .083 - .195 -.038 -.067 - . 043 -.295 25 - . 091 - .20 8 -.428 -.060 - . 041 -.029 -. 500 26 . 053 - • 0 61 -.2 58 .051 .204 -.141 . 012 27 .057 · . 391 - .123 .169 - .0 64 • 341 -. 229 28 .237 0 . 500 .098 0 .294 .169 29 . 044 .2 63 . 151 .056 .073 .107 - . 111 30 -.022 -.026 . 151 .114 - .0 22 - .0 80 -.238 31 -.0 35 -.0 75 - .134 -.2 83 .3 22 • 411 -.316 32 -.022 -.462 - .233 .17 5 .145 -.160 -.280 33 .200 - • 071 .428 .13 6 .267 .131 0 34 .0 93 .0 83 .1 67 . 316 0 - • 218 . 447 494 35 . 301 .122 .575 .689 .194 -.053 .098 36 . 2 71 .3 28 .25 8 .270 .121 .181 • 5 68 37 .17 5 .115 . 321 -.043 .182 - .057 . 4 81 38 -.099 -.220 - .1 67 -.045 -.009 .098 -.299 39 -.030 - . 012 - 0140 .17 2 - . 2 81 - .124 -.170 40 -.056 -.033 -.240 - .152 .121 0 - • 201 41 .005 -.247 .3 87 .026 .191 -.076 -.452 42 .045 .187 .33 8 .275 -.200 -.080 -.060 43 .164 -.060 .598 .206 . 331 -.008' .140 44 • 031 .187 .07 6 .037 .146 .133 .076 4.5 .170 .463 .338 . 210 .073 0 .408 46 -.070 .144 -.299 .042 .177 - .164 -.167 47 -.07 4 -.216 - • 319 0 -.256 -.096 .302 Variable 22 with variables: 23 .500 . 44 7 .177 .354 • 7 91 .804 .378 24 .17 7 .338 0 -.087 .15 7 .209 .190 25 .25 5 . 484 .15 6 0 - .122 .496 • 513 26 - .142 -.478 .0 86 .057 - . 310 - .167 -.060 27 0 - .182 ~.302 .259 . 310 .033 .022 28 -.028 .218 . 467 -.098 0 - • 218 .225 29 .3 46 . 441 .600 . 211 .242 .439 .289 30 • 217 .29 4 .600 . 271 .043 .070 .056 31 .287 .365 -.167 . 071 .278 .344 . 463 32 .191 .183 ·-.272 .0 68 -.049 . 3 61 .324 33 .306 .630 .040 .280 - .191 • 461 .422 34 .088 .28 4 .122 -.250 0 .333 .250 35 .50 2 1 . 00 .033 .3 44 .327 . 665 .756 36 . 455 .632 .224 .566 .244 • 381 • 601 37 .453 .517 .556 .632 -.079 .4 71 .460 38 . 402 .29 3 .433 . 410 .279 .180 .538 39 -.039 .149 -.015 .151 - .105 - . 313 . 010 40 .087 0 .161 .050 - • 015 0 - • 041 41 .205 . 258 . 312 -.008 . - .168 .170 .243 42 .239 .248 .480 .3 55 • 061 .070 .195 43 .377 . 65 2 . 458 .355 .0 32 .140 .577 44 .257 .408 .185 .17 5 . 435 .22 8 .437 45 - .003 - .192 - .1 61 .1 62 -.121 0 .333 46 .0 65 - .09 8 - .129 -.092 • 314 .295 .036 47 .101 .12 0 .185 0 .279 -.020 .084 495 Variable 23 with variables: 24 -.03 4 .066 -.356 .11 6 -.220 .185 -.265 25 -.079 -.086 -.256 -.115 .167 . 3 71 .365 26 -.115 -.265 -.295 .12 2 -.070 -.035 -.052 27 - .12 9 -.400 .122 .02 8 -.224 - .199 -.195 28 .060 - .194 .500 -.098 0 - .119 .354 29 .225 . 321 - .101 .038 .440 • 665 .0 84 30 .101 .302 - .101 .100 .138 .287 - . 013 31 . 219 .302 .624 .507 -.043 .239 .089 32 .007 -.459 - .171 .111 .105 .299 .092 33 .235 .435 .158 .098 .101 .5~4 -.067 34 -.058 0 . 431 -.250 -.707 0 - • 316 35 .397 .673 .325 .344 .346 .686 -.094 36 . 217 .350 .190 • 258 .405 .248 - .121 37 .209 .304 .275 - .106 .288 .5 86 - .140 38 .09 2 .118 -.285 .102 .15 6 .053 .244 39 -.246 -.038 -.35 3 -.225 - . 015 -.267 - • 471 40 -.049 .054 -.463 .23 4 .023 0 - .131 41 .079 -.024 . 014 .008 • 015 .140 .246 42 .183 .285 . 301 .022 .023 • 287 · ·.197 43 .273 . 5 68 .342 .022 .055 . 217 .330 44 .232 .285 . 2 71 - .180 .452 . 428 .123 45 .096 . 313 -.073 -.044 -.037 0 .269 46 .145 - .143 .112 .033 -.045 .379 .418 47 .095 .141 .229 0 .15 6 - .114 .080 Variable 24 with variables: 25 .236 .194 .586 .558 0 - .12 6 .252 26 -.084 -.638 .179 .200 -.397 .037 .052 27 -.057 -.267 0 -.267 • 419 - .186 . -.038 28 - . 312 -.577 -.632 -.488 0 .049 - . 051 29 .240 .294 .2 36 .10 7 .153 .064 . 411 30 . 292 .116 .236 .107 • 314 .141 . 5 84 31 . 236 .200 -.3 27 . 3 71 .255 • 091 .509 32 .3 80 .183 • 4 71 . 721 .204 -.052 • 514 33 .03 2 .539 -.360 -.023 -.023 - .187 .270 34 - . 031 .083 -.289 . 447 . 316 0 - .15 8 35 .12 7 .329 -.278 .15 8 .105 .279 .288 36 .085 .394 . 211 .169 - . 016 -.308 .118 37 .101 .840 -.050 • 051 - . 431 - .111 .066 38 .181 .378 .300 ·-. 211 .135 .114 .089 39 - . 019 .3 24 -.077 .009 .199 -.425 .237 40 .484 . 461 .678 .455 .513 0 .324 496 41 .229 .24 7 .0 97 .036 .233 .282 .324 42 -.001 .13 7 - . 3 87 -.302 -.054 .141 .150 43 - .107 .452 - . 514 -.386 - .162 .204 - . 218 44 -.025 .299 -.355 -.067 - .153 .llO • 218 45 -.344 - .154 -.342 -.302 -.324 0 -.584 46 .230 . 012 .021 .629 -.007 . 450 .134 47 .12 2 . 216 .277 0 .140 .439 - . 219 Variable 25 with variables: 26 .044 -.446 .320 .239 .325 - 0158 -.344 27 .033 - .124 .354 -.058 -.103 .182 .033 28 .088 .04 8 -.250 - .194 0 .185 .356 29 . 201 .079 .233 .055 .300 .203 . 241 30 . 2 71 .309 .233 .055 .320 .246 .167 31 .192 -.030 -.250 .134 .357 .289 .350 32 . 441 .577 .449 .574 .225 .380 .304 33 .057 .03 6 -.397 .222 -.060 .309 .088 34 -.188 -.278 -.408 -.200 - . 316 0 -.488 35 .17 4 .655 -.387 .500 -.237 .240 .452 36 .079 .289 - .149 -.083 .055 .169 -.058 37 .024 .108 - .187 .079 . ll9 .078 -.289 38 .2 45 .25 8 .340 -.0 62 .169 .316 .436 39 -.006 .022 - • 012 -.256 - .169 .086 . 218 40 .229 - . 014 . 483 .150 .250 0 .038 41 .182 . 271 - .02 4 .141 .14 7 .17 6 .258 42 .079 .208 - .146 - . 216 .089 .246 .205 43 - .070 . 091 -.283 -.260 .250 - .13 7 .230 44 -.022 -.022 .150 - .12 9 -.042 .131 .279 45 -.274 -.085 -.405 -.372 - . 281 0 -.038 46 .092 -.252 - . 012 .17 5 -.204 -.027 .496 47 -.057 .12 5 .150 0 - .123 -.032 -.344 Variable 26 with variables: 27 -.060 .124 - .169 - .134 -.394 .093 .109 28 .104 .289 - . 316 .167 0 -.347 .395 29 . 091 -.236 .300 .308 . -.124 . 017 .194 30 .052 -.236 .300 .308 - • 313 - . 017 .101 31 - .101 .030 - .189 0 0 -.240 -.356 32 .20 3 -.058 .214 .357 .302 .189 .149 33 .059 -.357 . 283 .378 - .13 8 -.042 .289 34 - .230 .030 -.708 .250 - .63 2 - . 218 -.250 35 .022 -.294 . .0 16 -.060 .02 6 -.076 .250 36 .Oll -.221 - • 016 -.059 .124 -.029 - • 081 497 37 .112 -.446 -.022 .255 .204 .2 89 .034 38 . 05 4 -.236 .332 .21 6 -.397 .250 .217 39 .100 - . 061 . 314 -.229 - .1 32 .053 .250 40 .17 5 -.243 . 311 .279 .146 0 .189 41 .13 5 - .1 92 .174 - .140 .2 23 .556 ~253 42 -.067 - .182 .069 .279 -.569 - . 017 -.096 43 -.079 -.577 • 111 -. 420 -.032 .200 · -.037 44 .027 -.23 6 .250 . 216 -.05 4 .042 .292 45 -.033 -.ll5 .189 - .183 .210 0 - .140 46 -.050 .058 -.295 .183 -.038 -.256 -.067 47 .0 42 - .13 2 .289 0 - • ll2 - .100 .253 Variable 27 with variables: 28 .303 • 671 - .15 8 . 356 0 .338 .395 29 -.090 - . 418 - .196 -.121 - .1 87 .028 - . 017 30 -.009 -.578 - .196 - .121 .083 .125 .242 31 -.095 -.603 - .102 - .100 .0 69 - . 051 .350 32 - . 013 -.342 • 214 - .121 0 .193 .10 8 3'3 - .177 - . 620 -.287 -.027 -.338 .125 .122 34 .032 .293 - . 091 0 .632 -.577 0 35 -. 020 -.080 -.238 .500 -.080 0 .19 6 36 -.164 - .12 6 -. 081. .12 6 -.567 -.055 - .139 37 -.2 58 - . 471 - . 315 .0 69 -.523 - . 231 -.258 38 -.086 -.472 -.248 .277 .10 8 .093 .009 39 -.081 -.288 -.304 .169 - . 201 - .187 .349 40 .023 -.456 .167 .036 -.033 0 . 427 41 - .117 -.362 - .122 -.239 - • 216 .189 .116 42 - .020 -. 472 -.440 -.083 .214 .125 . 217 43 - .17 9 - . 411 -.204 . 017 -.256 - .158 - . 122 44 -.003 -.045 -.079 .097 - . 061 .10 6 - .145 45 - . 001 -.258 - .13 8 .203 -.256 0 .194 46 - .038 - .169 . 085 -.388 • 061 - • 210 .191 47 - .130 -.248 -.079 0 - .149 - . 010 - .194 Variable 28 with variables: 29 .042 -.40 8 -.37 8 . 293 0 .0 26 .283 30 .0 69 -.577 -.37 8 . 293 0 .350 .267 31 - .142 - . 661 -.33 3 -.258 0 .0 69 0 32 - .16 2 - .3 86 -.750 -. 463 0 -.327 .5 48 33 .032 -.535 0 .327 0 .049 .333 34 .0 87 . 612 .500 0 0 -.730 -.333 35 .263 . 488 .167 .1 67 0 .026 .655 36 - .165 - .189 -.333 - .1 67 0 - .2 62 - .167 498 37 -.087 -.599 . 316 .039 0 - .109 -.333 38 -.075 -.683 .158 .17 4 0 - .12 5 .283 39 . 021 -.408 .500 .19 2 0 -.395 .386 · 40 - .184 -.775 - . 791 .0 98 0 0 .267 41 -.087 -.522 0 .17 4 0 -.347 .267 42 .075 -.577 .158 .258 0 .350 0 43 .165 -.258 .500 .683 0 .108 .225 44 .028 -.293 .500 -.522 0 .108 .283 45 . 016 0 .189 .098 0 0 .033 46 - .12 7 .192 0 -.488 0 -.406 .220 47 0 -.522 0 0 0 .233 .141 Variable 29 with variables: ' 31 .264 .408 -.289 .051 .430 .228 .386 32 .300 .189 -.187 • 25 6 .357 .107 .769 33 .307 .206 .181 -.007 .227 . 611 .664 34 - .15 4 - .15 8 -.320 - • 316 .060 0 - • 316 35 .3 25 .30 4 .200 .149 .139 .469 .553 36 .353 .596 .026 .13 2 .471 .302 • 511 37 .29 4 .332 0 • 005· .292 .432 .424 38 .3 57 .500 .366 .179 .133 .279 .722 39 .15 6 .333 .029 .359 - .133 -.048 .386 40 .423 .35 6 .495 .502 .357 0 .552 41 .2 85 .356 .282 .209 . 213 .148 . 412 42 . 5.13 . 612 .459 . 618 .441 • 418 .330 43 .159 . 312 .227 -.007 .209 .057 . 313 44 . 5 05 • 612 .199 .30 6 • 5 67 .786 .742 45 -.097 .19 6 .043 -.029 - .185 0 - .363 46 .080 - .167 - .101 .049 .209 .244 -.088 47 -.034 .167 - .169 0 .032 -.345 - .108 Variable 30 with variables: 31 .269 .272 -.289 .220 .3 68 -.043 .624 32 .275 .365 - .187 . 25 6 .144 -.080 .592 33 .121 .15 4 .181 -.007 .080 .180 .505 34 -.226 -.527 -.320 -.316 - .158 -.745 -.378 35 .0 93 .12 7 .200 .149 - . .026 .142 .17 5 36 .155 .258 .0 26 .187 .242 -.123 .179 37 .076 .017 0 .005 .0 62 .02 8 - . 0 61 38 .340 .592 .366 .202 .273 .503 .279 39 .151 .395 .02 9 .296 .133 .084 .189 40 . 551 .492 .495 .533 .506 0 • 812 41 .372 .492 .282 .222 .302 .260 .606 499 42 .702 . 846 .459 . 656 .626 1.00 . 486 43 .045 .167 .2 27 .036 .053 .070 .07 8 44 .11 6 .175 .199 .141 .075 .462 . 210 45 - .17 0 .206 .043 . 021 -.233 0 -.446 46 .13 2 -.099 - .101 .11 2 .194 .330 .046 47 -.049 .230 .169 0 - .079 -.20 7 -.208 Variable 31 with variables: 32 .258 .426 .333 .120 .150 -.354 .607 33 -.071 .452 .289 -.167 - .1 69 -.043 -.316 34 .297 .300 0 0 .258 .500 0 35 .106 .375 . 408 -.500 .284 .238 .258 36 .297 .535 . 667 . 417 .118 -.250 .167 37 .118 .468 .6 67 -.449 -.044 0 • 091 38 .180 .272 -.430 -.033 . 561 . 312 • 214 39 -.073 .13 4 -.222 - . 415 - . 081 -.071 - .134 40 . 391 .58 3 -.356 .507 .408 0 .624 41 .172 .784 - . 241 - . 312 0 0 . 418 42 . 061 0 .2 89 -.071 .039 -.043 . 418 43 .110 .302 • 241 -. 071 .284 .043 .039 44 .038 . 612 0 - • 461 .185 .228 .13 4 45 -.327 .068 - .134 - . 690 - • 5 67 0 0 46 .304 .354 .043 .690 .725 - . 091 .069 47 .12 9 0 .222 0 .094 0 .134 Variable 3 2 with variables: 33 . 010 .209 -.492 -.095 -.200 .306 .467 34 -.022 -.283 .120 .447 - .100 - • 218 -.316 35 .133 .320 - • 316 .267 -.080 .119 .645 36 . 151 .270 .37 8 -.236 .049 . 3 71 .203 37 .107 .191 - • 051 .086 .128 .224 .045 38 .198 • 3 65 -.204 - .055 -.050 .189 .523 39 & .066 .488 - .171 -.025 - . 201 .119 .2 43 40 . 473 .236 .685 . 310 .570 0 . 4 61 41 . 261 . 4 71 .171 -.093 .0 45 .472 .390 42 -.004 .189 -.328 .077 -.284 -.080 .31 2 43 -.083 0 - .171 -.422 .296 -.008 .141 44 -.030 .092 - . 514 . 010 .195 -.070 . 553 45 -.403 - .2 73 -.492 -.299 - .1 82 0 - . 4 61 46 .024 0 -.089 .359 - • 212 .052 .144 47 .078 . 218 .343 0 -.066 - .113 .149 500 Variable 33 with variables: 34 - . 091 .192 --.389 .577 0 - .149 .15 8 35 . 531 - . 564 . 575 .3 50 .3 54 .665 .485 36 .48 6 .342 .149 . 628 .599 .57 3 . 630 37 .50 3 .7 80 . 092 . 65 9 • 415 .5 32 .367 38 .295 . 347 .086 .343 .187 .051 .574 39 .19 6 .20 4 . 316 .2 80 .053 -.030 . 376 40 .002 . 289 - • 381 .118 -.007 0 . 3 71 41 . 097 .229 -.040 - .1 2 9 .05 3 .242 .29 5 42 .16 3 .13 9 .463 .007 .187 .180 .20 4 43 .339 . 7 91 .349 .118 .13 5 .109 .339 44 . 319 .139 .247 .2 33 .287 .368 • 516 45 .0 54 -.270 .690 .007 -.025 0 -.439 46 -.080 - .13 9 - • 219 - . 12 7 - . 017 . 064 - .124 47 - . 031 .15 8 -.334 0 .187 -.207 - • 013 Variable 34 with variables: 35 -.102 . 316 -.239 -.5 77 0 • 218 0 36 .115 .089 .3 65 0 -.378 -.250 . 447 37 .155 . 213 . 4 71 . 612 -.730 - .149 . 600 38 -.224 - . 617 - .0 55 0 .378 - • 25 8 -.25 8 39 - . 111 -.639 - . 068 .4 47 -.37 8 . 091 .293 40 -.077 -.284 - . 218 .316 - . 5 98 0 -.250 41 - . 019 -.192 .492 -1.00 - . 395 - . 218 -.250 42 - .119 -.527 -. 05 5 . 316 .632 -.745 - .250 43 - .117 . 192 .123 - • 316 • 316 -.655 -.500 44 - .1 81 .15 8 -. 480 .2 00 .350 0 - • 316 45 - . 081 - .184 - . 218 -.316 0 0 .378 46 .196 - . 051 .2 89 .632 .060 • 218 .316 47 . 018 - .19 2 .320 0 -.39 5 0 0 Variable 3 5 with variables : 36 . 351 .564 .160 . 488 , . 5 2 7 .3 60 .350 37 . 379 .555 .194 .15 8 .17 8 . 534 .218 38 .00 3 .094 -.433 - . 02 6 .113 -.070 • 4 71 39 - . 041 0 -.305 .149 .417 - . 462 . 351 40 - .07 9 - . 120 - .15 8 -. 026 .154 0 0 41 . 15 7 .189 .158 . 372 .105 .155 .250 42 .092 .15 2 . 484 - .2 44 -.207 .142 .171 43 . 298 . 679 . 316 .189 -.277 . 211 .316 44 . 341 .378 . 217 . 12 2 .367 . 25 1 .63 2 45 .11 8 0 .22 4 . 258 - .10 5 0 -.094 501 46 . 044 .189 -.255 -.25 8 .419 . 301 -.079 47 .088 .130 -.040 0 .105 .235 0 Variable 3 6 with variables: 37 . 669 • 616 .88 6 . 458 . 5 87 .529 .894 38 .317 . 390 • 3 21 .2 43 .30 8 .181 .373 39 .21 6 .30 2 - .12 7 .404 - .168 .13 7 .179 40 .180 .183 . 277 . 216 .105 0 .099 41 .2 39 .369 . 332 0 .1 68 .204 -.034 42 .077 .179 - .10 8 .150 .105 - .123 .302 43 .209 .52 2 .155 .105 .043 -.083 .2 87 44 .250 .72 4 - .307 .2 43 .480 .03 8 .373 45 . 081 -.101 - .15 8 • 216 .15 7 0 .17 4 46 -.040 - .179 - .370 .187 .390 - .151 -.302 47 .131 0 .171 0 .308 - . 316 .302 Variable 37 with variable s: 38 .2 58 .344 .3 27 .2 86 .079 -.030 .38 6 39 .10 9 .2 92 - .190 .247 -.187 - .1 61 • 314 40 .12 7 .340 .048 . 033 .037 0 .0 69 41 .171 .079 .370 - .1 83 . 2 71 .204 - .171 42 .0 84 .088 .1 92 .033 -.060 .028 -.171 43 .378 . 601 .408 . 356 .375 .232 • 211 44 .2 34 .344 -.073 .406 • 321 .19 6 .3 35 45 - .009 - • 313 - • .1 92 -.2 44 .078 0 . 071 46 - .143 -.062 -.408 -.098 -.038 - .186 -.3 44 47 .0 69 .18 9 . 218 0 .079 -.273 .102 Variable 38 with variables: 40 .307 .53 9 .279 . 316 .140 • 311 .260 41 . 327 .539 . 246 - .104 -.096 0 .4 74 42 .2 37 .422 . 02 8 . 215 . 3 82 .556 .1 00 43 .250 .386 .112 . 041 .0 67 . 503 .548 44 . 216 .422 .11 2 . 1 06 .009 .200 .697 45 - .0 73 0 .086 -.010 - • 20 l .067 -.066 46 -.ll 8 -.283 -.342 -.274 .3 00 0 - .123 47 .0 99 .253 .217 0 .253 .23 8 -.060 502 Variable 3 9 with variables: 40 .152 .365 - .12 5 .333 .190 - .100 .280 41 .107 .3 65 - . 071 0 .113 .155 .209 42 . 031 .296 -.0 28 .123 -.0 36 .084 - .121 43 .0 68 .175 .20 6 .123 - .14 7 - .121 .075 44 .101 • 296, .104 .365 -.375 - .10 6 . 215 45 .107 .078 .377 .12 0 -.023 0 - . 215 46 - .163 - . 2 61 - .1 62 - .118 - .147 -.084 - . 221 47 - . 019 .171 .027 0 - .141 -.369 - .178 503 APPENDIX E BIBLIOGRAPHIC SOURCES USED I N CODING WARFARE VARIABLES LISTED BY STANDARD SAI'1PLE I DENTITY NUMBER OF EACH SOCIETY 1 . Schapera , I saac 1930 The Khoisan People s of South Afri ca : Bushmen and Hottentots . London : Routledge and Kegan Paul. 2 . Marshall , Lorna 1961 Sharing , Talking, and Giving : Relief of Social Tens i ons Among !Kung Bushmen . Africa 31( 3): 231-249 . 1960 !Kung Bushman Bands . Africa 30(4 ): 325- 355 . 1965 The !Kung Bushmen of the Kalahari Desert . I n Peoples of Africa . J . L . Gibbs, Jr . , Ed . New York : Holt, Rinehart and Winston . pp . 241 - 278 . Schapera, I saac 1930 The Khoisan Peoples of South Africa : Bushmen and Hottentots . London : Routledge and Kegan Paul. Thomas , Elizabeth Marshall The Harml ess People. New York : Alfred Knopf . 3 - Junod , Henri A. 1927 The Life of a South Afri can Tribe , Second revised edition . London : Macmi l lan . 4 . Gluckman, Max 1 941 Economy of the Central Barotse Plain . 1 951 The Lozi of Barotseland in North Western Rhodes i a . I n Seven Tribes of British Central Africa . Elizabeth Colson and Max Gluckman , Eds . London : Oxford Uni vers ity Press . pp . 1 - 93. Turner , V. W. 1952 The Lozi People of North- Western Rhodesia . London : Haze l , Watson a nJ. Viney , Ltd . 504 5 . Childs , G. M. 1949 Umbundu Ki nshi p and Characte r . London : I nter- national African I nst i tute . Edwards , A. D. 1962 The Ovimbundu Under Two Sovere i gnt ies. London : Oxford University Press . Hambly, W. D. 1934 The Ovi mbundu of Angola . Chicago : Field Museum of Natural His tory . McCulloch , M. 1952 The Ovimbundu of Angola . London : I nter- nat i onal African I nstitute . 6 . Ward, Herbert 1969 Five Years With the Congo Ca nnibals , Thi rd edi tion. New York : Negro University Press . Weeks , John H. 1914 Among the Primi t i ve Bakongo . London : Seeley , Service and Co., Ltd . Wing , J . V. 1959 ~tudes Bakongo , Deuxieme edi ti on . Be l gique : Desclee de Brouwer . 7 . Smith, E . W. and A. M. Dale 1968 The I la- Speaking Pe ople s of Northern Rhodesia . New York : Unive rsity Books . 8 .. Wilson , Monica 1 95 1 Good Company . London : Oxford University Press . 1957 Rituals of Kinshi p Among the Nyakyusa . London : Oxford University Pre ss . 9 . Bleek , D. F . 1930 The Hadzapi or Watindiga . Africa 4 : 273- 285 . Woodburn , ,James 1968 An I ntroduction to Hadza Ecology . I n Man the Hunter . R. B. Lee and I . DeVore , Eds . Chicago : Aldine . pp . 49- 55 . 10 . Be i de l man , T. 0~ 1967 The Matrilineal Peoples of Eastern Tanzania . London : I nternationa l African I nsti tut e. 197 1 The Kaguru : A Matrilineal People of East Africa . New York : Holt , Rinehart and Winston .. Chr i stensen , J . B. · 1963 Utani : Joking, Sexual License and Social Obli- gations Among the Lu.guru . Ameri can Anthropol- ogi st 65 :1 314-1 327 . Scheerder and Ta s tevin 1950 Les Wa l u guru . Anthropos 45 : 24 1-286 . 505 Yotmg , Ro l and and H. Fosbrooke 1960 Smoke I n the Hills. Evanston : Northwestern University Press . 11. Kenyatta , Jorno 1953 Faci ng Mount Kenya . Lo ndon : Mart i n Seker and Warburg . Lambert , H. E . 1956 Kikuyu Social and Political Ins titut ions . London : Oxford University Press . Lawren , W. L . 1968 Masai and Kikuyu : an Historical Analysis of Culture Transmi ssi on . Jour nal of African History 9(4) : 571 - 583 . Leakey , L. S . B. 1952 Mau Mau and the Kikuyu. London : Methuen . Mi ddleton , John and Greet Kershaw 1965 The Kikuyu and Kamba of Kenya . London : Internat iona l Afri can Institute . Mungeam , G. H. 1970 Masai and Kikuyu Responses to the Establ ish- ment of British Admini stration . Journal of African Histor y 11 (1) : 127-1 43 . Routledge, W. S. and K. Routledge 1910 With a Prehistoric People . London : Edward Arnold . 12 . Mai r , Lucy P. 1934 An African People i n the Twentieth Century . London : G. Routledge and Sons . Ros coe , John 1911 The Baganda : An Account of Their Native Cus- toms and Beliefs . London : Ma cmillan . 13. Turnbull, Colin Mo 1 96 1 The Forest People . New York : Simon and Schuster . 1965 The Mbut i Pygmies : An Ethnographic Survey . Anthropological Papers of the American Museum of Natural History 50(3) . 14 . Brown, H.F . , 1944 The Nkurnu of t J~e Tumba . Africa 14 : 431 - 446 . Hul staert , G. 1938 Le Mariage des Nkund6 . Memoirs de l ' I nstitut Royal Colonial Belge 8 :1 - 520 . 506 15 . Dugast , I . 1955 Monographi e de la Tribu des Ndi ki . Traveaux et Memoires de l 'Inst itut d ' Ethnol ogie 58 : 1-824 . McCul loch , M., M. Li ttlewood , I . Dugast 1954 Peoples of the Central Cameroons . London . 16 . Bohannon , Laura 1958 Political Aspe cts of Tiv Social Organi zation . I n Tribe s Without Rulers . J . Mi ddlet on and D. Ta it , Eds . LoLdon : Routle dge and Kegan Paul . pp . 33- 66 . Bohannon , Laura and Paul Bohannon 1953 The Ti v of Central Nigeri a . London : Inter- nat iona l African I nst itute . Bohannon , Paul and Laura Bohannon 1968 Tiv Economy . Evanston : Northwestern Uni- vers ity Press . Downes , R. M. 1933 The Tiv Tribe . Kaduna : The Government Printer . East , Rupert 1939 Akiga ' s Stor y . London : Oxford University Press . Sahli ns , Marshal l 1961 The Segmentary Lineage : An Organizat i on of Predatory Expansion . American Anthropologist 63 : 322- 345 . 1 7 . Basden , G. T. 1921 Among the I bos of Nigeri a . Lo ndon : Seeley , Se rvi ce and Co., Ltd . 1966 Ni ger I bos . LonQon : Frank Cass and Co. Jeffreys , M. D. w. 1956 I bo Warfare . Man (June) : 77- 79 . Uchendu , Vi ctor c. 1966 The I gbo of Southeast Ni geria . New York: Holt , Rine hart and Winston . 18 . Argyl e , w. J. 1966 The Fon of Dahomey . Oxf ord : Clarendon Press . Lloyd , Alan 1964 The Drums of Ku.mas i: the St ory of the Ashanti Wars . London : Longmans Green and Co. Rattray, R. S . 1 91 6 Ashant i Proverbs . Oxford : Clarendon Press . 1929 Ashanti Law and Constitution . Oxford : Clarendon Press . 507 Tordo:ff , William 1965 Ashant i Under the Prempehs , 1888- 1935 . London : Oxford Universit y Press . 20 . Little , Kenneth L. 1951 The Mende of Sierra Leone . London : Rout ledge and Kegan Paul . 21. Ames , D. W. 1959 Wolof Co- operat i ve Wo1·k Groups . In Continuity and Change i n African Cultures . W. R. Bascom and M. J . Kerskovi ts , Eds . Uni versity of Chi cago Pre ss . pp . 224- 237 . Forde , C. Daryll and P . M. Kaberr y 1967 West African Ki ngdoms i n the Ni neteenth Cen- tury . London : I nternational African I ns titute .. Gai ley, Harry A. 1965 A History of the Gambia . New Yor k : Praeger . Gamble , D. P . 1957 The Wolof of Senegambi a . London : Inter- national African I nstitute . Qui nn , Charlotte A. 1972 Mandingo Kingdoms of t he Senega~bia . Evans ton : Northwestern University Press . 22 . Monte il, c. 1927 Le s Bambara du Segou et du Kaarta . Pari s : La Rose . Pacques , V. 1954 Les Bambara . Pari s : Presses Universitaires de France . 23 . Fortes , Meyer 1967 The Dynamics of Clanship Amo ng t he Tall ens i . London: Oxford University Pres s . 24 . Miner , Horace 1953 The Primitive Ci ty of Timbuctoo . Princeton University Press . 25 . Stenni ng, Derrick J. 1959 Savannah Nomads . London : I nternational African I nst i tute . 1965 The Pastoral Fulani of Nort hern Nigeri a . I n Peoples of Africa . J . L . Gibbs , Jr . , Ed . New York : Holt, Rinehart and Winston . pp . 363- 40 1 . 508 26 . Smi th , Michael Garfield 1960 Government i n Zazzau . Lo ndon : I nternational African Institut e . 1965 The Hausa of Northern Ni ger i a . In Peoples of Africa . J . L . Gi bbs , Jr . , Ed . New York : Ho lt, Rinehart and Winston . pp . 11 9-1 55 . Tri mingham , J. S. 1962 A History of I slam i n West Africa . Londpn : Oxford Univers i ty Press . 27 . Barth, Henry 1859 Travels and Discoveries in North and Central Africa . New York : Ha rpe r and Bros . Brenner , Louis 1973 The Shehus of Kukawa . Oxford : Clarendon Press . Cohen, Ronald 1966 The Dynamics of Feudali sm in Bornu . Boston Unive rsity Publications on Africa 2 : 87- 105. 1967 The Kanuri of Bornu . New York : Holt, Ri nehart and Winston . Cohen , Ronal d and John Mi ddl eton , Eds . 1970 Fr om Tribe to Nat ion in Africa . Scranton : Chandler Publi shing Co. 28 . Evans- Pritchard, E . E. 1957 Zande Border Raids . Africa 27 : 217- 231 . 1957 Zande Warfare . Anthropos 52 : 239- 262 . 1971 The Azande : Hi story and Political Institu- tions . Oxford : Clarendon Press. 29 . Nachtigal , Gustav 197 1 Wadai and Da rfur . Vol. IV, Sahara and Sudan . A. G. B . Fisher and H. J . Fish er , Trans l ators . Berke l ey : Un iversity of California Press . 0 1 Fahey, R . S. 1973 Slavery and the Slave Trade i n Dar Fur . Journal of African Hi s tory 14(1) : 29- 43 . 30 . Nade l, S . F. 1947 The Nuba . Lo ndon : Oxf ord University Press . 31 . Evans- Pritchard , E . E . 1948 The Divine Ki ngshi p of the Shilluk of the Nilot i c Sudan . Cambri dge University Press . Howell , P . P . 194 1 The Shilluk Sett leruent . Sudan Notes and Records 24 : 47-6 6 . 1952 Observations on the Shilluk of the Uppe r Nile . Africa 22 : 97-11 9 . 509 Mercer , Patricia 1 97 1 Shilluk Trade and Politics From the Mid-1 7th Cent ury to 186 1 . Journal of African Hi s tory 12(3) : 407- 426 . 32. Seligr.an , Charl es G. and Brenda Z. Seligman 1932 Pagan Tribes of the Nilot ic Sudan . London : Routledge and Sons, Ltd . 33 . Bieber, F . J . 1920- Kaffa . Munster . [Whi l e this source is i n 1923 German and has not been used for this study , it i s the basic source .--VWN] 34 . Forde , C. Daryll 1949 Habitat, Eco nomy , and Society . London : Methuen and Co . , Ltd . Fosbrook e , H. A. 1948 An Administrative Survey of the Masai Social System . Tanganyika Notes and Records 26 : 1- 50 . Huntingford , G. W. B. 1953 The Southern Ni l o- Hamites . London : I nter- national African I nstitute . Kenyatta , Jomo 1953 Facing Mount Kenya . London : Martin Seker and Warburg , Ltd . 35 . Hallpike , C. R. 1972 The Konso of Ethiopia . Oxford : Clarendon Press. 36 . Lewis, loan M. 1961 A Pastoral Democracy . London : Oxford Univer- sity Press . 196 1 Force and Fission i n Northern Soma li Lineage Structure. American Anthropologi st 63 : 94-11 2 . 1965 The Northern Pastoral Somali of the Horn . I n Peoples of Africa . J . L . Gibbs , Jr ., Ed . New York : Holt , Rinehart and Wins ton . pp . 319-360 . 37 . Messing , S . D. 1957 The H:Lghl and- Plateau Amha ra of Eth iopia . Ph . D. Dissertat ion , Univers ity of Pennsyl- vania . 38 . Paul, A. 1 971 A History of the Beja Tribes of the Sudan . London : Frank Cass and Co. , Ltd . 510 39 . Callender . Charles and Fadwa el Gui ndi 197 1 , Life Cr i sis .Hi tuals Amo ng the Kenuz . Case Western Reserve University Studies i n Anthro - pology 3. Fernea , Robert A. 1973 Nubians i n Egypt : Peaceful Pe opl e . Aust i n : Univers i ty of Texas Press . Geiser , Peter 1973 The Myth of the Dam . American Anthropologi st 7 5 ( 1 ) : 18 4- 19 4 . 40 . Briggs , Ll oyd Cabot 1958 The Living Races of t he Sahara . Papers of the Peabody Museum of Archaeo l ogy and Ethnol- ogy, Harvard Univers i ty , 28(2 ). 1960 Tri bes of t he Sahara . Cambridge : Harvard Uni vers i ty Press . Chapelle , Jean 1957 Nomades Noi rs du Sahara . Pari s : Librairi e Plon . 41 . Lhote , Henri 1955 Le s Touaregs Du Hoggar . Paris : Payot . Nicola i sen , Johannes 1963 Ecology and Culture of the Pastoral Tuareg . The Nat i onal Mus eum of Copenhagen . Ro dd , Frar1cis P . 1926 Peopl e of the Ve il. · Lond on : Macmillan . 42 . Coon , Carl eton S . 193 1 Tribes of the Rif . Harvard African Studi es 9 . Hart , Davi d M. 1970 Clan , Lineage , Lo ca l Communi ty and the Feud i n a Ri fian Tribe . I n Peopl es and Cultures of t he Mi ddle East , Vol . 2 . Louise E. Sweet , Ed . New York : Na tura l Hi story Press . Woolman, Davi d S. 1968 Rebel s in the Rif . Stanford University Press . 43 . Breasted , James H. 1905 A History of Egypt from the Earli est Times to the Persian Conquest . New York : Scri bner ' s Sons . Cottrel, Leonard 1969 The Warri or Pharaohs . New York : G. P . Putnam ' s Sons . 44 . Devaux , Roland 196 1 Ancient Israel . Lond on : Darton , Longman and ~L1o dd . 51 1 Holy Bibl e Books of DeuteroEomy and Jo shua . Old Te sta- ment . 45 . De l aporte , L. 1970 Mesopotami a . Trans . by V. Gordon Childe . London : Rout ledge and Kegan Paul . Mcquee n , James G. 1964 Babylon . New York : Frederick A. Praeger . Saggs, H. W. F . 1966 The Greatness That Was Babylon . New York : Hawthorn Books . 46 . Musil, Alois 1928 The Manners and Customs of the Rwala Bedouins . New York . 47. Lerner , Daniel 1958 The Passing of Tradit i onal Society . New York : Free Pres s . Makal , Mahmoud 1954 A Village i n Anatolia . London : Vallentine Press . Pierce, J . E . 1964 Life in a Turkish Village . New York : Holt, Rinehart and Wins ton . Stirling , P . 1965 Turkish Village . Lo ndon . 48 . Coon , Carleton S. 1950 The Mountain of Gi ants . Papers of the Peabody Museum of American Ar chae ology and Ethnology 23(3) :1-1 05 . Durham , M. E . 1910 High Albani a and Its Customs i n 1908 . Journal of the Royal Ant hro po logica l Ins titute XL (July-De c) : 453- 472 . 1928 Some Triba l Or i g ins , Laws , a nd Cust oms of the Balkans . London . Whitaker , I an 1968 Tribal St ructure and National Politic s i n Albani a , 1910-1 950 . In History and Socia l Anthropology . I . M. Levi s , Ed . London : Tavi stock Publica tions . pp . 253-293 . 49. Brand , C. E . 1968 Roman Military Law . Austin : Uni versity of Texas Press . 512 Carcopino , Jerome 1940 Daily Life i n Ancien t Rome . Translate d by E. 0 . Lorimer . New Haven : Yale Universi t y Press . Fri edlander , Luuwig 1965 Roman Life and Manners Under t he Early Etiipire. New York : Barnes and Noble . Millar , Fergus 1967 The Roman Empire and It s Neighbors . New York : Del a corte Pre ss . Salmon , E . T. 1966 A History of the Roman Worl d , 30 B.C.-A.D. 138 . London : Methuen . Watson, G. R. 1969 The Roman Soldier . Ithaca : Corne ll University Press. 50 . Medhurst , Kenneth 1972 The Basques . London : Minority Rights Group. Thomas , Hugh 1 961 Tbe Spanish Civil War . New York : Harper and Bros . 51. Arensberg , Conrad and Sol on T. Kimball 1968 Family and Communi ty i n Ireland , Second edition . Cambridge : Harvard University Press . Coogan , T. P. 1966 I re l and Since the Ri sing . New York: Praeger . 1970 The I . R. A. New York : Praege r. 52 . Pehrson , R. N. 1950 Culture Cont a ct Wi thout Conflict in Lapl and . Man 50 :1 57-160 . Pelto, Pert ti 1962 I ndi v i dualism I n Skolt Lapp Society . Kansatietellinen Arkisto 16 . Helsinki : Finnish Antiquit ies Society . Vorren, ¢rnulv and Ernst Manker 1962 Lapp Life and Customs . London : Oxford Univer- sity Press . 53 . Donner , Kai 1954 Amo ng the Samoyed In Siber i a . New Haven : HRAF , Inc. Kopytoff , I. 1955 The Samoyed . New Haven : HRAF , Inc. 54 . Dunn , S . P . and E . Dunn 1967 The Peasants of Central Rus s ia. New York : Holt , Rinehart and Winston . 513 Fains od , M. 1963 How Russia i s Ruled . Cambridge : Harvard University Press . 55 . Allen , w. E . D. and Paul M-:J.ratoff 1953 Caucas i an Battlefields . Cambri dge University Press . 56 . Alle n , w. E . D. and Paul Muratoff 1953 Caucasian Battlefields . Cambr i dge University Press . Bli ss , Rev . Edwin Munsell 1896 Turkey and the Armenian Atrbcities. New York : Edgewood Publ ishing Co . Nalbandian , Louise 1963 The Armenian Revol utionary Movement . Berkeley : University of Ca lifornia Press . 57 . Barth , Fredrik 1954 Father ' s Brother's Daughter Marriage in Kurdistan . Southwestern Journal of Anthro- pology 10:1 64-1 71 . Leach , E . R. 1940 Social and Economic Organization of the Rowanduz Kurds . London School of Economics Monographs in Social Anthropology 3:1-74 . 58 . Barth , F . K. 196 1 Nomads of South Persia . London : George Allen and Unwin. 59 . Callard, K. 1957 Pakistan : A Political Study. New York : Macmillan . Callard , K. and R. S . Wteeler 1963 Pakistan . I n Major Governments in Asia . G. McT . Kahin , Ed . Ithaca : Cornell University Press . pp . 419- 532 . Eglar , Z. S. 1960 A Punjabi Village in Pakistan . New York : Columbia University Press . 60 . Grigson , W. V. 1938 The Maria Gonds of Bastar. I ndian Bran ch : Oxford University Press . 61 . Rivers , W. H. R. 1906 The Todas . London : Macmillan . 514 62 . Culshm-, , W. J . 1949 Tribal Heritage . London : Lutterworth Pre ss . Orans, Mart in 1956 The Santal . Detroit : Wayne State University Press. Cohn , Bernard S . 1955 The Changi ng Status of a Depressed Caste . American Anthropologi cal Assoc i ation MEL1oir 83 : 53- 77 . 1 971 I ndia : The Social Anthropo l ogy of a Civi li- zation . Englewood Cliffs : Prent i ce- Hall. Cornell University , Department of Far Eastern Studies , I ndia Program 1956 Uttar Pradesh : An Area Handbook . New Haven : HRAF , Ind . Opler , M. E . and R. D. Singh 1954 The Division of Labor i n an I ndian Village . In Reader i n General Anthropology . C. S . Coon , Ed . New York : H. Holt . pp . 464- 496 . Palmer , Norman D. 1963 India . I n Major Governments in Asia . G. McT. Kahin , Ee["° Ithaca : Cornell Universi ty Press . 64 . Durand, A. G. A. 1899 The Making of a Frontier . London : J . Murray . Lorimer, D. L . R. 1935 The Burushaski Language . Os lo. Lorimer , E . 0. 1938 The Burusho of Hunza . Antiqui ty 12 : 5- 15 . 1939 Language Hunting in the Karakoram . London : George . Hudson , Alfred E . 1938 Kazak Social Structure . Yale University Pub- lications in Anthropol ogy 20 . M-i.;_rdock , G. p . 1934 The Kazaks of Central Asia . In Our Primit i ve Contemporaries . New York : Macmi l lan . pp. 135-1 62. 66 . Rupen , R. A. 1956 General Chara cter of the Society . In Mongolian People t s Republic . W. M. Ballis , Ed . New Haven . Vol . 1 , pp . 1- 48 . 1964 Mongols of the 20th Century . Bloomington : Indi ana Uni versity . 1966 The Mongolian People ' s Republic . Stanford University ~ Hoover Institution Studies . 515 Vreeland , H. H. 1954 Mongol Community and Kinship Structure . lfo1·r Haven : HRAF , I nc. 67 . Lin , Yueh- hwa 1947 The Lolo of Liang Shan . Shanghai : Comm . Press . 68 . Gorer , Geoffrey 1938 Himalayan Village . London : M. Joseph . Morris, John 1938 Living with Lepchas . London : W. Heinemann . 69 . Burling, Robbins 1963 Rengsanggri . Philadelphia : University of Pennsylvania Press . 70 . Hutton, John H. 192 1 The Serna Nagas . London : Macmillan . 71 . Nash , Manning 1965 The Golden Road to Modernity . New York : John Wiley and Sons . Silverstein , Josef 1964 Burma. I n Major Governments in Southeas t Asia . G. McT . Kahi n , Ed . Ithaca : Corne ll University Press . pp . 75- 182 . 72 . Leach , Edmund R. 1970 Politica l Systems of Highland Burma. London School of Economics Monographs on Social Anthropology 44. Milne , Mary Lewis 1924 The Home of an Eastern Clan : a Study of the Palaungs of the Shan States . Oxford : Clarendon Press . 73 . Buttinger , Joseph 1967 Vietnam : A Dragon Embattled . New York : Praeger . Kahin, G. McT ., Ed . 1964 Goverrunents and Politics of Southeast Asia . Ithaca : Cornell Univers i ty Press . Tanham, George K~ 196 1 Communist Revolutionary Warfare . New York : Praege r . 74 . American University, Washington, D. C. 1966 The Rhade. In Minari ty G-roups in the Republic of Vietnam . Department of the Army Pamphlet No . 550- 105 : 651 - 718 . 5 16 Donoghue , John D. 1963 The Rhade of South Viet Nam. Current Anthro - pology 4(4) : 382- 384 . Mole, Robert L. 1970 The Montagnard.s of South Vietnam . Rutland , Vermont : Charle s E . Tuttle Co . 75 . Aymonier , E . 1901 Le Cambodge . Paris. Coedes , G. 1968 The Indianized States of Southeast Asia . Honolulu : East - West Center Press. Wales , H. G. 1952 Ancient South- East Asi a n Warfare . London : Be rnard Quarit c h , Ltd . 76 . I nsor , D. 1963 Thailand , A Polit ical , So cia l, and Economi c Analysis. New York : George All en and Unwin . Wilson , D. A. 1964 Thailand . In Governments and Politics of Southeast Asia. G. :McT . Kahin , Ed . Itha ca : Cornell Univers ity Press . pp . 3- 74 . 7 '7 I • Dentan , Robert K. 1968 The Semai : A Nonviolent People of Malaya . New York : Holt , Rinehart and Winst on . 78 . Man, E . H. 1932 The Nicobar I slands and Their People . Guilford : Billing and Sons . Whitehead , George 1924 In the Nic obar I slands . London : Seeley, Service and Co . 79 . Man , E . H. 1932 On the Aborigi nal I nhabitants of the Andaman I slands . London : Royal Anthropologi cal I nsti t ute of Great Britain and Ireland . Radcliffe-Brown , A. R. 1922 The Andaman I slanders . Cambridge Universi ty Press . 80 . Ba iley, J . 1963 An Account of the Wild Veddahs of Ceylon . Transactions of the Ethnological Society of London 2 : 278-320 . Seligmann , C . G . and B. ,z. Seligmann 1911 The Veddas . Cambridge University Press . 517 8 1 . Linton , Ralph 1933 The Tana la . Field Museum of Natural Hi story Anthropologica l Series 22 : 1- 334 . 82 . Gullick , J . M. 1 958 Indi genous Politica l Sys tems of Western Malaya . London School of Economics Monographs on Social Anthropology 17 : 1- 151 . 83 . Anderson , Benedict R. 1972 Java in a Time of Revol ut ion : Occupat ion and Resistance 1944- 1946 . I thaca : Cornell University Press . Kahin , G. McT . 1952 Nationali sm and Revo luti on i n I ndonesia . I tha ca : Cornell University Press . Kosut , Hal 1967 I ndones i a : The Sokarno Years . New York : Facts on File . 84 . Be lo, Jane , Ed . 1970 Tradit ional Balinese Culture . New York : Columbi a University Press . Covarrubias , Miguel 1942 The I s l and of Bali . New York : Alfred A. Knopf . 85 . Fr eeman , J . D. 1955 Repor t on the Iban of Sarawak . Kuching : Government Pr i nt i ng Offi ce . Gomes , Edwin 1911 Seventeen Years Among t he Sea Dyaks of Borneo . London : Seeley . Low , Hugh 1848 Sarawak , I t s I nhabitants and Productions . London : Richard Bent ley . Roth , H. Lung , Ed . 1892 The Native s of Borne o . Journal of t he Royal Anthropologica l I ns titute 21 : 110- 137 ; 22 : 22- 64 . 86 . Nimmo, H. A. 1968 Reflect i ons on Badjau History . Philippine Studies 16 : 32- 59 . 1972 The Sea People of t he Sulu . San Francisco : Chandler Publishing Co . 87 . Wa l es , Horace G. Q. 1952 Ancient South- Eas t As i an Warfare . London : B. Quaritch . 518 88 . Hueting , A. 1921 De Tobeloreezen in hun de nken e n doe.n . Bi,j dr agen tot de Taal-, Land- , en Volkenkunde 77 : 217-385; 78 : 137- 342 . Rie del , J . G. F . 1885 Galela und Tobeloresen . Zeitschrift fi.ir Ethnologie 17 : 58- 89 . [These are the primary s ources for the Tobelorese ; they were not used i n this study , however . The codings are left as blanks--VWN . ] 89 . DuBois, Cora 1944 The Peopl e of Alor . Minneapolis: University of Minnesota Press . 90 . Hart , C. W. M., and A. R. Pilling 1860 The Tiwi of North Australia . New York : Holt , Ri nehart and Winston . 91 . Basedow , H. 1925 The Australi an Aborigina l . Ade laide : F. W. Preece. Spencer , B. and F . J . Gillen 1927 The Arunta . London : Macmillan . 92 . Reay, Marie 19 53 Social Control 11..rnongst t he Oro kai va . Oceania 24 : 110-11 8 . Williams , Francis E. 1930 Orokaiva Society . London : Oxford University Press . 93 . Serpenti , L. M. 1965 Cultivators i n t he Swamps . Assen : Von Gorcum. 94 . Pospisil, Leopold 1958 Kapauku Papuans and Their Law . Yale Uni ver- sity Publications i n Anthropol ogy 54 . 1963 The Kapauku Papuans of West New Gui nea . New York : Holt , Rine hart and Winston . 95 . Whiting , J . W. M. 1941 Becoming a Kwoma . New Haven : Yale University Pre s s . Whiting , J . W. M. and S . W. Reed 1938 Kwoma Culture . Oceania 9 :1 70-216 . 96 . Mead , Margaret 1930 Growing Up i n New Guinea . New Yor k . 519 196 1 The Manus of the Admi ralty I slands . I n Coo pe r a tion and Compet i t i on Arnong Primi t ive Peopl es . M. Me ad , Ed . Bosto n : Beac on . pp . 2 10- 239 . 97 . Powdermake r , Hortense 1931 Vi tal Statist ic s i n New Irela nd . Human Biology 3 : 35 1-375 . 19 32 Feasts i n New lreland . Ameri can Anthropolo- gist 34 : 236-247 . 1933 Life In Le s u. New York : W. W. Norton . 98 . Malinowski , Broni s l aw 1920 War and Weapons Among t he Nat i ves of the Trobriand I s l ands . Man 20 : 10- 12 . 1922 Arg onauts of the We stern Pacific . Lo.nd on : Routledge and Sons . 1926 Crime and Cu stom i n Savage So c iety . New York : Harcourt Brace . 99 . Oli ver , D. L . 1955 A Solomon I s l and Society . Ca mbridge Univer- sity Press . 100 . Firth, Raymond 1936 We, t he Tikopia . London . 1960 Succession to Chi eftai nshi p i n Tik op i a . Oceania 30 : 161-1 80 . 1961 History a n d Traditions of Tikopia . Wellington : The Polyne s i an So ciety . 1 01 . Lane , R. B. 1965 The Melane s i ans of South Pentecost . I n Gods , Gho sts and Men i n Melanesia . P . La wrence and M. G. Meggit t , Eds . Melbourne : Oxford Un i- vers ity Press . pp . 250- 279 . 102 . Williams , T. 1858 Fiji and t he Fi j i ans . London : Alexander Heylin . 103 . Leenhardt , Maurice 1930 Notes d ' ethnologie s n~o-ca l ~donienne . Traveaux et Memori es de o ' I nstit u t d ' Ethno l ogi e 8 : 1- 340 . Oliver , D. 1 . 1 96 1 The Pacific I slands . New York : Doubleday . Saussol, Alain 1 97 1 New Cale doni a : Colon ization and React i on . I n Land 1l'enure i n the Pacific . Ron Crocombe , Ed . Me l bourne : Oxfo rd Un ivers ity Press . pp . 22 7- 24 7 . 520 104 . Vayda , Andrew 1960 Maori Warfare . New Zea land Polynesian Society Mao r i Monographs 2 . 105 . Handy , E . S. C. 1932 The Native Culture of the Marquesas . Bulle - tins of t he Bernice P . Bishop Museum 9 : 1- 358. Linton, Ralph 1939 Marquesan Culture . In The Individual and His Society . Abram Kardiner, Ed . New York : Columbia University Press . pp . 138- 196 . 106 . Murdock, G. P . 1934 Our Primi tive Contemporarie s . New York : Macmillan . Stair , J . B. 1897 Old Samoa . London : The Religious Tract Society. Turner , G. 1884 Samoa . London : Macmillan . 107 . Lambert , B. 1966 .Ambilineal Descent Groups in the Northern Gilbert I slands . .American Anthropo logist 68 : 64 1- 664 . 1970 The Gilbert Islands . I n Land Tenure i n the Pacific . R . G. Crocombe , Ed . Melbourne : Oxford University Press . 1 08 . Fins ch , Ott o 1881 Kriegfuhrung auf den Marshall-Inseln . Gartenlaube 29 : 700-703 . Kramer , A. and H. Nevermann . 1938 Ralik- Ratak . Ergebnisse der Sudsee - Expedition 1908-1 0 . G. Thilen iusj Ed . Hamburg . [ HR.AF translation used . Snefft, A. 1903 Die Marshall-Insulaner . Rechtsverhaltnisse von eingeborenen Vo l kern in Afrika und Ozeanien pp . 425-4 55 - Berli n . [Trans l ated by HRAF . _7, Spoehr , A. 1949 Majuro : A Village in the Marshall Islands . Fieldiana : Anthropology 39 : 1-266 . 109. Bol]ig , Laurent ius 1927 Die Bewohner der Truk-Inseln . Munster : Aschendorffsche Verlagsbuch l andlung . [Trans - lated i n the HRAF . ] 521 Goodenough , W. H. 1 95 1 Property , Kin , and Corrnnuni ty on Tru:L Yale University Publications in Anthropology 46 : 1 - 1 92 . Hall , E . T. and K. J . Pelzev 1946 The Economy of the Truk I slaLds . Honolulu : U. S . Commercial Co . Kramer , Augustin 1932 Truk . Hamburg : Friederichsen . 11 0. Muller , Wi l helm 1 917 Yap . Hamburg : Friederichsen . Senfft , Arno 190'( Ethnographische Be i trage uber die Karolineni nsel Yap . Petermanns Mitteilungen 49 : 49- 60 , 83- 87 . Tetens , Alfred and Johann Kubary 1873 Die Carolineni nsel Yap oder Guap nach den Mittheilungen von Alf . Tetens and Johann Kubary . Hamburg : Museum Godeffroy Journal 1 : 84-1 30 . [Transl ated in the HRAF .] 111 . Barnett , Homer G. 1960 Be i ng a Palauan . New York : Holt, Rinehart and Winston . Force , Roland W. 1960 Leadership and Cultural Change i n Palau. Fieldiana : Anthropology 50 : 1-21 1. Force , Roland and Maryanne Force 1972 Just One House . Bernice P . Bishop Museum Bullet i n 235 . 112. Barton , R. F. 1 91 9 Ifugao Law . Berkeley : University of Cali- fo rnia Press . 1930 The Half- Way Sun . New York : Brewer and Warren . 11 3 . Mabuchi, T. 1960 The Aboriginal Peoples of Formosa . I n Social Structure in Southeast Asia . G. P . Murdock , Ed . Viking Fund Publications in Anthropology 29 :1 27-1 40. · 11 4 . Fei , H. 1946 Peasant Life in China . New York : Linebarger , P . M. A., D. Chu , and A. W. Burks . · 1956 Far Eastern Governments and Politics : Chi na and Japan. Princeton Universi ty Press . 522 11 5 . Shirokogoroff , S . M. 1924 Socia l Organi zat ion of the Mancnus . Royal Asia tic Society, North China Branch , Extra Vol. 3 : 1-1 94 . 11 6 . Osgood , Cornelius 195 1 The Koreans and Their Culture . New York : The Ronald Press. 11 7 . Beardsley , R. K., J . W. Hall, a nd R. E . Wa rd 19:":>9 Village J apan . Linebarger , P . M. A., D. Chu , and A. W. Burks . 1956 Far Eastern Governments and Politics : China and Japan . Princeton University Press . 11 8 . Batche lor, J . 1895 The Ainu of Japan . New York . 1927 Ainu Life and Lore . Tokyo. Takakura , Shin 1 ichiro 1960 The Ainu of Northern Japan . Philadelphia : American Philosophical Soci ety . 119 . Shternberg , L. Semya i rod u narodov severo- vostochnoi Azi i Leningrad . 120 . Jochelson , Wa l demar 1926 The Yukaghir and Yukaghirized Tungus . Memoirs of the American Mus eum of Natural History 13 : 1-4 69 . 121 . Bogoras , W. 1904- The Chukchee . Memoirs of the American Museum 1909 of Natural Hi s tory 11 : 1-703 . 122 . Osgoo d , Cornelius 1958 I ngalik Socia l Culture . Yale Uni versity Pub- lications i n Anthropology 53 :1-289 . 1959 Ingalik Men t a l Culture . Yale University Pub- lications in Anthropology 56 :1-1 95 . 123 . Colljns, Henry Bascom, Jr. 1947 The I s l ands and Thei r Peo ple. Ir- The Aleutian I slands : The i r Pe ople and Natural History . H. B. Collins , J r ., Austin H. Clark , and Egbert H. Walker , Eds . Wa shi ngton , D. C.: Smiths onian I nst itution. pp . 1-30 , 72-74.' Veniaminov , I van E. P . 1040 Zapi ski ob ostrovakh unalashki nskago otdela . St . Petersburg : Izdano Izhdiveniem Ros s iysko- Ameri kansko i Kompanii . [Translated in the HRAF . ] 523 124 . Je nness , Di amond 1922 The Life of the Coppe r Eski mo . Ottawa : F . A. Acland . 125 . Hallowe ll, A. I. 1940 Aggression i n Saul teaux Society . Psychi at ry 3 : 395- 407. Lane , Kenneth 1952 The Montagnais I ndians , 1600- 1640 . Kroeber Anthropolog ica l Soci ety Pape rs 7 :1-62 . Leacock , Eleanor 19 54 The Montagnai s "Hunt i ng Territory " and t he Fur Trade . American Anthropolog ica l Associ a tion Me mo i r 78 . Speck , F . G. 193 5 Naskapi . Notman : Unive rsi ty of Ok l ahoma Pre ss. Thwaite s , R. G. , Ed . 1906 Jesuit Relat ions and Allie d Documents . New York : Pageant Book Co . 126 . Bo c k , Philip K. 1966 The Micmac I ndi ans of Re st i gouche : Histor y and Co ntemporary Description . Nat iona l Museum of Canada Bullet i n 213 , Anthropologica l Seri e s No . 77 . Walli s , Wilson D. and Ruth Sawte ll Wallis 1955 The Micmac I ndians of Eastern Ca nada . Minneapolis : University of Minnesota Press . 127 . Dunn i ng , R . W. 1959 Social and Economic Change Among the Ojibwa . Toronto . Ha llowell, A. I. 1940 Aggression i n Saulteaux Society . Psy c h i atry 3 : 395- 407 . 1955 The Nor thern Oj i bwa . I n Cultur e and Experi- e nce . University of Philadel phia Press~ pp . 112-1 24 . 128 . Helm , June 19G1 Th e Lynx Poin t Peopl e . Bullet i n of the National J,1useum of Canada 176 : 1-1 93 . Honigmann , J . J. 1946 Et hnography and Acculturation of the Fort Nelson Sl ave . Yale Univers ity Publi c a t i ons i n Ant hropology 33 :1-1 69 . Ma cNeish , June He l m 1956 Leadership Among t he Northeast Athabascans . Anthropo logica 2 : 131-1 63 . 524 129 . Hcni gmann , J . J . 1949 Culture and Ethos of Ka ska Soci e ty . Yale Ur:i - versity Publicat ions in Anthropology 40 : 1-368 . 1954 The Kaska I ndians . Yale Uni vers ity Publica- tions i n Ant hropology 51 : 1- 163 . 130 . Birket- Smith, K. a nd F . de Laguna 1938 The Eyak I ndians of the Coppe r River Delta . Kpbe nhavn . 1 31 . Curti s , E . S. 1 916 The North Ameri can I ndi an XI : 115-1 75 , 186-1 93 . Norwood . Drucker , Philip 1965 Cul tures of the North Pacifi c Coast . San Francisco: Chandler Publishi ng Co . Murdock , G . p . 1934 Ha i da . I n Our Pri mitive Contemporaries . New York : Macmillan . pp . 221-263 . Swantou , J . R. 1909 Contributions to the Ethnology of the Haida . Memori es of the American Museum of Natura l History 8 : 300 . 132 . Mcllwraith , T. F . 1948 The Be lla Coola I ndians . Toronto . 133 . Elmendorf , W. W. 1960 The Structure of Twana Cul ture . Research Studie s , Monographic Supplement No . 2 . Washington State Uni versity . 134 . Kroeber, A. L. ·1945 A Yurok War Reminis cence . Southwestern J our- nal of Anthropology 1 : 31 8-332 . 1960 Comparat i ve Notes o n t he Structure of Yurok Cult ure . I n The Structure of Twana Cul ture . W. W. Elmendorf . Research Studies , Mono- graphic Supplement No . 2 . Washi ngton State Univers ity . 1 35 . Agi nsky , B. W. arid E . G. .Aginsky 1967 Deep Valley . New York : Stein and Day. Barre t t , S. A. 1952 Materi a l Aspects of Pomo Culture . Bulletin of the Public Museum of the City of Mi l waukee , v . 26 , pt . I . Gifford , E . W. 1926 Clea r Lake Pomo Soci e ty. Uni versity of Ca lifornia Publica tions i n American Archae- ology and Et hnology 18 (2) : 287- 390 . 525 Kroeber , A. L . 1953 The Yu.rok . · In Handbook of the Indians of California . Berkeley : California Book Co . pp . 1- 97 . Loe b , E . M. 1926 Pomo Fol kways . University of Ca liforni a Pub- licat ions in American Archaeology and Ethnology 19 : 183-1 85 , 200- 212 . 1 36 . Gayton , A. H. 1948 Yokuts and Western Mono Ethnography . Anthro- pological Records 10 : 1-301 . Kroeber , A. L. 1953 The Yokuts. In Handbook of the I ndi ans of California . Berkeley : Cali forni a Book Co . pp . 474- 543 . 137 . Blyth, B. 1938 Northern Paiute Bands i n Oregon . American Anthropologi st 40 : 402- 405 . Stewart , Omer 1941 Northern Paiute . Anthropological Recor~s 4 : 361-446 . Wheeler- Voegelin , E . 1955- The Northern Paiute of Central Oregon . Ethno- 1956 history 2: 95-132 , 241 - 272 ; 3 : 1-10 . Whit i ng, Beatrice 1950 Paiute Sorcery. Viking Fund Publications in Anthropology 15 : 1-110. 138 . Spier , Leslie 1930 Klamath Ethnography . University of Ca l iforni a Publications in Ameri can Archaeology and Ethnology 30 . Stern , Theodore , 1966 The Klamath Tribe . Seattle : Uni versity of Washington Press . 139 . Turney-High , H. H. 1 94 1 Ethnography of the Kutenai. . Memoirs of the American Anthropological Asso ciation 56 : 1- 202 . 140 . Cooper , J . M. 1956 The Gros Ventres of Montana . Catho lic Univer- sity of America Anthropological Series 16 : 1- 500 . Flannery, R. 1953 The Gros Ventres of Montana . Catholic Univer- s ity of America Anthropological Series 15 : 1-234 . 526 1 4 1 • Bowers , A. W. 1965 Hidatsa Social and Ceremonial Organization . Bul let i n of the Bureau of American Ethnology 194 : 1-528 . 142 . Dorsey, George A. 1904 Tradit ions of the Skidi Pawnee . London : Houghton , Mifflin and Co . 1906 Pawnee War Tales . American Anthropologist 8 : 337- 345 . Dorsey , George A. and J . R. Muri e 1940 Notes on Skidi Pawnee Society . Fi e l d Museum of Natural History Ant hropological Se r ies 27 ( 2 ) : 6 5-1 1 9 . Ho l der , Preston 1970 The Hoe and the Horse on the Plains . Lincol n : University of Nebraska Press . Murie, James R. 1914 Pawnee I ndian Societies . Anthropological Papers of the American Museum of Natural History 9 (pt . 7) . 1;le l tf i sh , Gene 1965 The Lost Uni verse . New York : Basi c Books . 143 . Dorsey , J . 0 . 1882 Omaha Sociology . Annual Reports of the Bur eau of American Ethnol ogy 3 : 205- 370 . Fletcher , A. C. and F . L . Flesche 1906 The Omaha Tribe . Annual Reports of the Bureau of American Ethno logy 27 :1 7-654 . Fortune , Rea 1932 Omaha Secret So ciet i es . Columbi a Uni vers i ty Contributions of Anthropology 14 :1-1 93 . 144 . Trigger , Bruce G. 1969 The Huron : Farmers of the North . New York : Holt, Rinehart and Winston . 145 . Swanton , J . R. 1928 Social Organization and Socia l Usages of the Creek Confederacy. Annual Report s of the Bureau of American Ethnology 42 : 23- 472 , 859- 900 . 1946 The I ndians of the Southeastern United States. Bulletin of the Bureau of Ameri can Ethnology 137 : 1-943 . 147 . Wallace , Ernest and E . Adamson Hoebel 1952 The Comanches : Lords of the South Plains . Norman : Uni versity of Oklahoma Press . 527 148 . Opler , Morris E . 194 1 An Apache Life Way . University of Chicago Press . 149 . Bunzel , Ruth 1930 Introductiofi to Zuni Ceremonialism . Annual Reports of the Bureau of Ameri can Ethnology 47 : 467- 544 . Smith , W. and J . M. Roberts 1954 Some Aspe cts of Zuni Law and Legal Proce dure . Plateau 17 :1- 5 . Stevenson , M. C. 1904 The Zuni I ndians . Annual Reports of the Bureau of American Ethnology 23 :1 3- 608 . 150 . Spier, Leslie 1928 Havasupai Ethnography. Anthropol ogica l Papers of the American Museum of Natural History 29 : 81 - 408 . 151 . Underhill , Ruth 1936 Autobiography of a Papago Woman . Me rrioirs of the American Anthropologica l Associa tion 46 : 1-64 . 1939 Social Organization of the Papago I ndians . Columbia Uni versity Contribut ions to Anthro- pology 30 :1-280. 1946 Papago Indian Re ligion . New York . 152 . Lumholtz , Carl 1903 ULknown Mexico . London: Macmillan . 153 . Gillmore , Frances 1964- The Ki ng Danced in the Marketplace. Tuc son : University of Arizona Press . Leon- Port i lla , Miguel 1963 Aztec Thought and Culture . Norman : University of Oklahoma Press . Vai l l ant , George 1950 The Aztecs of Mexico . Baltimore : Pelican Books . 1 54 . Foster , George M. 1943 The Geographical and Li nguist ic and Cultural Position of the Popoluca of Veracruz . American A...~thropologist 4-5 : 531-546 . 1942 A Primitive Mexican Economy . Monographs of the American Ethnological Society 5 : 1-115 . Seattle : Universi t y of Washi ngton Press . 528 155 . Roys, Ralph L. 1943 The I ndian Background of Colonial Yucatan . Washington : Carnegie Institut ion . Thompson , J . Eric S . 1966 The Rise and Fall of Maya Civilizat ion . Norman : University of Oklahoma Press . 156 . Conzemius , ·E . 1932 Ethnographi c Survey of the Miski to and Su.mu I ndians . Bulletin of the Bureau of Ameri can Et hnol ogy 106 : 1-191. Kirchhoff , P . 1948 The Caribbean Lowland Tribe s . Bulletin of the Bureau of American Ethnology 143 : iv, 219- 229 . 157 . Gabb , W. M. 1875 On the Indian Tribes and Languages of Costa Rica . Proceedings of t he American Philos ophi- ca l Society 14 : 463- 602 . Johnson , Frederick 1948 The Caribbean Lowland Tribes : the Ta lam an ca Divis i on . Bulletin of the Bureau of American Ethnology 142(4) : 247 . Stone , D. Z. 1963 The Ta l amancan Tr i bes of Costa Ri ca . Papers of the Peabody Museum , Harvard University 43 ( 2) : 1 - 1 08. 158 . Johnson , Frederick 1948 Central American Cultures . Bulletin of the Bureau of American Ethnology 142 (4) : 50 . 159 . Armstrong , J . M. a nd A. Metraux 1948 The Goajiro. Bulletin of the Bureau of American Ethnology 143 (4) : 360-383 . Gutierrez , de Pineda, V. 1948 Organizacion social en l a Gua ·ira . Revi sta del I ns tituto Et nologico Nacional 3(1) : 1-255 . Wi l bert , Johannes 1958 Kinship and So cia l Organization of the Ye Kuana and Goajiro . Southwestern Journa l of Anthropology 14 : 51- 60 . · 160 . Leyburn, J . G. 1941 The Haitian Peopl e . New Haven : Yale Univer- s i ty Press . 161 . Breton , R. and A. la Paix 1929 Relation de 1 1 i le de la Guadeloupe . I n Les Caraibe s , La Guadeloupe , 1635-1 656 . Joseph Hennard , Ed . pp . 45- 74 . Paris : Li brairie 529 Generale et Internationale . Du Tertre , Jean Baptiste 1667 Histoire generale des Anti lles habit e s par l es Francois . Pari s : T. I olly . Rouse , I . 1948 The Carib . Bullet i n of the Bureau of Ameri can Ethnol ogy 143 (4) : 547- 565 . Taylor, D. 1938 The Caribs of Dominica . Bul letin of the Bureau of Ameri can Ethnolocy 119 : 103- 159 . 162 . Wilbert , J ohannes 1970 Folk Literature of the Warao I ndi ans . Los Ange le s : Latin American Center . 1972 Survivors of El Dorado . New York : Praeger . 163 . Chagnon , Napole on 1968 Yanomamo : The Fierce People . New York : Holt , Rinehart and WiLlston . 164 . Gillin , J . 1936 The Barama River Cari bs . Papers of t he Peabody Museum , Harvard Uni versity 14( 2) : 1-274 . Rouse, I. 1948 Bulletin of the Bureau of American Ethnology 143 ( 4 ) : 547- 565 . 165 . Kahn , Morton C. 193 1 Djuka : The Bush Negroes of Dutch Gui ana . New York : Vi king Press . 166 . Murphy , Robert F . 1957 I ntergroup Host i l i ty and Social Cohesion . American Anthropologi st 59 :1 018-1 035 . 167 . Goldman , Irving 1948 Tribe s of the Vaupes-Caquetta Region . Bulletin of the Bureau of American Et hnology 143 ( 3) :763- '798 . 1963 The Cubeo . Illinois Studies in Anthropology 2 : 1-31 3 . 168 . Barrett , S . A. 1925 The Cayapa I ndians . I ndian totes and Mono- graphs 40 : 1- 476 . Murra , John 1948 The Cayapa and Color.ado . Bulletin of the Bureau of American Ethnology 143(4) : 277- 278 . 530 169 . Karsten , Rafael 1923 Blood Revenge , War , and Vict ory Feasts Among the Jibaro I ndi ans of Easte r n Ecuador . Bul - letin of t he Bureau of American Ethnology 79 . St i rling , Matthew W. 1938 Historical and Ethnographical Materi als on the J i varo I ndi ans . Bulletin of tne Bureau of American Ethnology 11 7 : 1- 148 . 170 . Carneiro , R. L . 1964 Shifting Cultivati on Among t he Amahua ca . Ni edersachsisches Landesmuseu.m , Volkerkind- liche Abhandlungen 1 : 9- 18 . Quote d by Donal d Lathrap , Subsistence and Ecology . I n Peopl es and Cultures of South America . Dani el R. Gross , Ed . Garden City : Doubleday . 1973 . 171 . Bram , Joseph 194 1 An Anal ysis of I nca Militari sm . Monograph of t he Ameri can Ethno l ogi ca l Socie ty . New York : J . J . Augustin . Rowe , J . H. 1946 I nca Culture at t he Time of t he Conques t . Bulletin of t he Bureau of American Ethnology 143 ( 2) : 183- 330 . 172 . LaBarr e , Weston 1948 The Aymara I ndi a ns of t he Lake Tit ica ca Pla - teau , Bolivia . Memoirs of t he American Anthropolog i cal Associat ion 68 . Ts chopik , Harry , Jr . 1946 The Aymara . Bulletin of t he Bureau of Ameri - can Ethno l ogy 143 ( 2 ) : 501-573 . 1 95 1 The Aymara of Chucuito , Peru . Anthropologi ca l Papers of the American Museum of Nat ural His- tory 44 : 133-308 . 173 . Holmberg , A. R . . 1950 Nomads of the Long Bow . Publicat ions of the I nstitute of Socia l Ant hropology , Smi t hsonian I nstitution 10 :1-104 . 174 . I.evl-Strauss , Claude 1948 The Nambicuara . Bulletin of the Bureau of American Ethnology 143(3) : 361-369 . 1972 Tri s te s Tropiq_ue s . Trans l ated from the French by John Russell . New York : Atheneum . 531 1 75 . Murphy , Robert F . and Buell Quain 1955 The Trumai Indians of Central Brazil . Mono- graph of the 11..i11eri can Ethno l ogica l Society 24 : 1 - 1 08 . 176 . Crocker , W. H. 196 1 The Cane l a Since Nimuendaju . Anthropo l ogical Quarterly 34 :69- 84 . Nimuendaju , Curt 1946 The Eastern Timbira . Uni versity of Californi a Publications in American Archaeol ogy and Ethnology 41 : 1-357 . 177 . Abbeville , Claude d ' 1614 Histoire de la mi ssion des Pe res Cap~cins en l ' I sle de Maragnan et terres cir convoi s i nes . Pari s : Francois Ruby . Metraux , A. 1948 The Tupinamba . Bulletin of the Bureau of American Ethnology 143(3) : 95- 133 . Souza , G. Soares de 1851 Tratado descript i ve do Brizil em 1587 . I nstituto Historic o e Geographico do Brazil, Revista 14 : 1-423 . Staden , H. 1928 The True Story of His Captivi ty . M. Lotts , Ed . London : George Routledge and So ns . Theve t, A. 1878 Le s singularitez de l a France antarctique . P . Gaffarel, Ed . Paris : Mai sonneuve . 178 . Metraux , A. 1946 The Botocudo . Bulletin of t he Bureau of American Ethnol ogy 143 (1 ) : 53 1-540 . Ni muendaj u , Curt 1946 Social Organization and Beliefs of the Botundo of East e rn Brazil . Southwestern Journal of Anthropology 2 : 93- 11 5. 179 . Maybury- Lewis , David 1965 The Savage and the I nnoce nt . Cleveland : World Publishing Co . 1967 Akwe- Shavante Society . London : Oxford Univers i ty Pre ~s . 180 . Henry , Jul es 194 1 The Jungle Peopl e . New York : Vintage Books 532 181 . Watson , J . B. 1954 Cayua Culture Change . Memoirs of the American Anthropological Association 73 :1-1 44 . 182 . Grubb , W. B. 1913 An Unknown People in an Unknown Land . London : See l ey , Service and Co . Metraux, A. 1946 Ethnography of the Chaco . Bulletin of the Bureau of American Ethnology 143( 1) : 197- 370 . 183 . Dobrizhoffer , M. 1822 An Account of the Abipones . London : J. Murray . 184 . Lat charn, Richard E . 1909 Ethnology of the Araucanos . Journal of the Royal Anthropological Institute 39 : 334-370 . Padden , Robert C. 1957 Cul tural Change and Mili tary Resistance i n Araucani an Chile . Southwestern Journal of Anthropology 13: 103-121. Ti tiev , M. 1 951 Araucanian Culture in Transit i on . Occasional Contribu.tions from the Museum of Anthropology , Univers ity of Michigan 15 : 1- 164 . 185 . Cooper , John M. 1946 Patagonian a nd Pampean Hunters . Bul letin of the Bureau of American Ethnology 143(1) : 127- 1 68 . Musters , G. C. 1873 At Home with the Patagonians . London : J . Murray . Viedma , A. de 1837 Des cripcion de l a costa meridional del sur . Coleccion de obras y documentos relatives a l a histor ia antigua y moderna de la provincias del Rio de l a Plata 6 : 63- 81 . Buenos Aires . 186 . Cooper , John M. 1946 The Yahgan . Bulletin of the Bureau of Ameri - can Ethnology 143(1 ) : 81-106 . 533 REFERENCES CITED 534 REFERENCES CITED Aberle, David F . 1968 General Discuss ion . In War : the Anthropology of Armed Conflict and Aggression . Morton Fried , J\farvi n Harris, and Robert Murphy , Eds . Garden City, New York : Nat ural History Press . pp . 97-1 00 . Allport , Gordon W. 1968 The Role of Expectancy . I n War , Revised edition . Leon Bramson and G. W. Goethals , Eds . New York : Basi c Books . pp . 177- 194 . Andreski , Stanisl av 1968 Mili tary Organization and Society . Berkeley : University of California Press . Anonymous 1972 I nnat ely Ambivalent . The Times Lit erary Supplement , March 3: 238 . Ardrey, Robert 1 961 African Genesis . New York : Atheneum . 1966 The Territorial I mperative . New York : Atheneum . 1970 The Social Contract . New York : Athenemn . Ayers , M. R. 1968 The Refutation of Determinism . London : Methuen . Baier , Kurt 1970 Responsibility and Ac tion . I n The Nature of Human Action. Myles Brand , Ed . Glenview, Illinois : Scott , Fore sman . Barnett , H. G. 1953 Innovation : the Basis of Cultural Change . New York : McGraw- Hill . Barraclough, Geoffrey 1967 An Intro duction to Cont emporary History . Bal~imore : Pelican Books . 1970 Deus le Volt? New York Review of Books 14(10) : 12-17 . 535 Benedi ct , Ruth 1959 The Natural History of War . In An Anthropologi st at Work . Margaret Mead . Boston : Houghton Miffl in . pp . 369- 382 . Benfer , Robert A. 1968 The Desirabil:Lt y of Small Sampl es f or Anthropolog i- ca l Inference . Ameri can Anthropo l ogi st 70 : 949- 95 1 . Berkowi tz , Leonard 1968 I mpul se , Aggressi on , and the Gun . Psychology To day . 1969 Roots of Aggressi on . New York : Athe rto n Press . Berreman , Geral d D. 1972 I s Ethnoscience Re l evant ? I n Culture and Cogni- ti on . J ame s P . Spradl ey, Ed-.- San Francis co : Chandl er . Bidney , David 1967 Theoreti ca l Ant hropo l ogy . New York : Sch ock en Books . Bohannon , Paul 1954 The Mi grat ion and Expansion of the Tiv . Afr ica 24 : 2- 16 . 1963 Soc i a l Anthropo l ogy . New York : Holt , Rinehart and Wi nston . oulding , Kenneth E . 1968 Am I a Man or a Mouse--or Both? I n Man and Aggr es- s i on , First editi on . M. F . Ashley Montagu , Ed . New York : Oxford Uni versity Press . pp . 83- 90 . Bramson , Leon and G. W. Goethal s , Eds . 1968 War : Studies from Psychology , Socio l ogy, Ant hropol- ogy . New York : Basi c Books . Bro ch , T. and J. Ga l tung 1966 Belli gerence Among the Pri mit i ves . J ourna l of Peace Research 1 : 33- 45 . Carne i ro , Robert L . 1972 A Theory of the Origi n of the State . I n Readi ngs i n An t hropology . J. J ennings and E . A. Hoebel, Eds . New York : McGraw- Hill. pp . 424- 432 . Carrighar , Sally 1968 War Is Not i n Ou.r Genes . I n Man and Aggressi on , First edition . M. F_ Ashl ey Montagu , Ed . New York : Oxford University Press . pp . 51- 58 . 536 Chagnon , Napo l e on 1968a Yanomamo Socia l Organizat i on and Warfa1·e . I n Har : the Ant hropology of Armed Conflict and Aggres0ion . Morton Fried , Marvin Harri s , a nd_ Ro' ert Murphy , Eds . Garden Ci ty , New York : Natural History Press . pp . 109- 159 . 1968b Yanomamo : the Fie rce People . New York : Holt , Rine - hart and Winston . Chaney , Richard P. 1970 Conceptual Cont ention : a Reply . American Anthropol- ogi st 72 : 1456- 146 l . 1972 On Scientifi c I nquiry and Sociocultural Da t a Pat- te rning . I n Models i n Archaeology . David Clarke , Ed . London : Methuen . 19,3 Comparative Analys i s and Ret roductive Reas oni ng : Conclusions i n Search of a Premi se . AA 75 : 1358- 1 375 • I Chaney , Richard P . and Roge lio Ruiz Revilla 1969 Sampling Methods and I nterpretat ion of Correlation: a Con~arat ive Analysis of Seve n Cross- Cultural Samples. Ameri can Anthropol ogi st 71 : 597-633 . Ccdy , Arthur B. 196 7 Can A Single Action Have Many Different Descri.p- tions? I nquiry 10 : 164-1 80 . Colby , B. N. 1966 Ethnographic Semant ic s : A Prelimi nary Sur vey . Cur - rent Anthropo logy 7(1 ) : 3- 32 . Coll i ns , Paul W. 1966 Comments [ on Driver 1966 ] . Current Ant hropology 7(2) : 149 . Co nklin , Harold C. 1955 Hanunoo Color Categories . Southwestern Journal of Anthropology 11 : 339- 344 . Cook , John'tl . · n . d . Relativi sm . Ufipub . Ms . Uni versity of Oregon . Crook , Johr.. H. ·1968 The Nature and Funct ion of Terri tor i a l Aggre s ion . I n Man and Aggression, Fj_rst edi tion . M. F . Ashl ey Montagu , Ed . New York : Oxf ord Uni versity Press . pp . 141- 17 8 . 537 D' Andrade, Roy G. n . d . Cultural Constructions of Reality . Unpub . Ms . San Diego : University of Californi a . Deming , Barbara n . d. On Revolut i on and Equilibrium . Palo Alto : Institute for the Study of Nonviolence . Dentan, Robert K. 1968 The Semai : A Nonviolent People of Malaya . New York : Holt, Rinehart and Winston. Diamond , Stanley 1968 War and the Dissociated Personality . I n War : the Anthropology of Armed Confl ict and Aggress i on . Merton Fried , Marvin Harris , and Robert Murphy , Eds . Garden City , New York : Natural Hi story Press . pp . 183-188 . Divale , William T. 1970 An Explanation for Tribal Warfare . Paper presented at the 69th Annual Meeting of the American Anthropo- logical Association , San Di ego . 19 71 Warfare in Primi t i ve Societies : a Selected Bi bliog- raphy . California State Col ~ege , Los Angeles : Center for the Study of Armament and Disarmament . 1973 Temporal Focus and Random Error i n Cross cul tural Hypothe s is Tests . Paper presented at the 72nd Annual Meeting of the American Anthropological Associ at i on , New Orleans . DDuglas , Mary 'i 966 Purity and Danger . New York : Praeger . D1neyfus , Hubert 1973 Artificial I ntelligence . Paper present ed at Cali- fo rnia State Univers i ty , Sacramento . Driver , Harold 19G5 Survey of Nvn1erical Cl assif i cat ion in Anthropol ogy . I n The Use of Computers in Anthropology . De l l Hymes, Ed . The Hague : Mouton . pp . 324-325 . 1966 Geographical- Historical Versus Psycho-Functional Explanat ions of Kin Avoidances . Current Anthro- pology 7 : 131-182 . Driver , Harold and Richard P . Chaney 1970 Cross- Cultural Sampling ana Galton ' s Problem . I n A Handbook of Method in Cult ur a l Anthropology . Garden City, New York : Natural History Press . pp. 990- 1003 . 538 Driver , Harold E . and William C. Massey 1957 Cor.1p2.rat i ve St udies of North American I ndi ans . T.cansactions of the American Philosophical Society 47 : 165- 456 . Driver , Haro ld E . anel Karl F . Schuessler 1967 Corre l ational Analys i s of Murdock ' s 1957 Ethno- graphic Sampl e . Ar:ierican Anthropo l ogist 69 : 332- 352 . Eggan , Fred 1966 })erspect i ves of the Study of Social Change . Chicago : Al dine . Ember , Me l vin and Carol R. Ember 1971 Conditions Favoring Matrilocal Ve rsus Patri local Re s i dence . Ameri can Antnropologi st 73 : 571-594 . Firth , Raymond 1959 Social Change i L Tikopia . New York : Macmillan. :Fison , Lo r imer and A. Howitt 1880 Kamilaroi and Kurnai : Group Marri age , Re l at ionship and Marri age by Elopement . New York : Humanities Press . Fox , Rob i n 1969 Re vi ew of War : the Anthropology of Armed Conflict and Aggression . M. Fried, M. Harris, and R. Murphy , Eds . American Anthropologist 71 : 314- 315 . Frake , Charl es 19 61 The Diagnosis of Di sease Among the Subanum of Mi ndanao . .American .Anthropologist 63 : 113-1 32. Freeman , Derek 1964 Human Aggressi on in .Anthropologica.l Perspective . I n The Natural History of Aggression . J. D. eart hy and F . J . Ebling , Eds . New York : .Academic Press . pp. 1 09-1 1 9 . Freeman, Linton 1965 Elementary .Applied Sta ti.st ies . New York : John Wiley a nd Sons . Freilich, Morris 1966 Comments [on Driver 1966 ] . Current Anthropo logy 7(2) :1 53-1 54 . .Frie d , Morton 1967 'l'he Evolution of Political Society . New York : Random House . 539 Fried, Morton , Marvin Harri s , and Robert Murphy , Eds . 1968 War : the Anthropology of Aggression ana Armed Con- flict . Garden City, New York : Natural Hi story Press . Gamow , George 1962 Gravity . Garden City , New York : Anchor Books , Doubleday . Gilula , Marshall F . and David N. Daniels 1969 Violence and Man 1 s Struggle to Adapt . Science 164 (April 25) : 396 - 405 . Gorer , Geoffrey 1968 Man Has No "Killer " Instirict. I n Man and Aggres- sion, First edition . M. F . Ashley Montagu , Ed . New York : Oxford University Press . pp . 27- 36 . Hans on , Norwood Russell 1958 Patterns of Discovery . Cambridge : at the University Press . Harris, Maiovin 1968 The Rise of Anthropol ogica l Theory . New York : Thomas Y. Crowell . 1 97 1 Culture , Man , and Nature . New York : Thomas Y. Crowe ll. 1972 Warfare Old and New . Natural History 81 (3) : 18- 20 . Heider , Karl 1970 A Papuan Culture in the Highl ands of West New Guinea . Viking Fund Publicat i ons i n Anthropology 49 . Chi cago : Aldine. Hempel, Carl G. 1959 The Logic of Functional Analysis . In Symposium of Sociologica l Theory . Llewellyn Gross , Ed . New York : Harper and Row . Hobhouse, L. T. , G. C. Wheeler, and M. Ginsburg 1915 The Material Culture and Social I nstitutions of the Simpler Peoples. London : Chapman and Hall. Hoebel , E . Adamson 19 58 Man in the Primitive World . New York : McGraw-Hill . Holloway, RaJ~h L., Jr . 1968 Human Aegression : the Need For a Species- Specific Framework . l:q War : the Anthropology of Armed Con- flict and Aggression . M. Fried , M. Harris , and R. l\forphy , Eds . Garden City, New York : Natural 540 History Press . pp . 29-48 . 1969 Culture : A Human Domain . Ci.;;.rrent Anthropology 10 (4) , pt . II : 395- 412 . I nst i tute for Social Research 1971 Newsletter 1(10) : 1-6 . Ann Arbor : University of Michigan . Jorgensen , Joseph G. 1966 Geographical Clusterings and Functional Explanat i ons of I n- Law Avoi dances : An Anal ys i s of Comparative Method. Current Anthropology 7:1 61-1 69 . Kaplan , David 1960 Law of Cultural Dominance . I n Evolution and Cul- ture . Marshall D. Sahlins and E. R. Service , Eds . Ann Arbor : University of Michigan Press . pp . 69- 82 . 1968 The Formal- Substant i ve Controversy i n Economic Anthropology : Reflections on I ts Wider I mplications . Southwestern Journal of Anthropol ogy 24 : 228-251. Katz, J . and J. Fodor 1963 The Structure of a Semant ic Theory . Language 39 : 170- 210 . Kobben , A~ J . F . 1967 Why Exceptions ? The Logic of Cross-Cultural Analy- sis . Current Anthropo l ogy 8 : 3- 34 . Le a ch , Edmund 1968 Don 1 t Say "Boo " to a Goose . I n J.VIan and Aggression , First edition . M. F . Asnley Montagu , Ed . New York : Oxford University Press . Leba r , Frank M. 1970 Codi ng Ethnographic Materials . I n A Handbook Of Method in Cultural Anthropology . · Raoul Naro l l and Ro nald Cohen , Eds. Garden Ci ty , New York : Natural Histor y Press . Lee , Bi.chard B. 1968 What Hunters Do For a Living , or, How to Make Out On Sca rce Resources . I n Man the Hunter . R. B. Lee and Irve n DeVore, Eds . -Tihicago : Aldine . pp . 30- 48 . Leon-Portilla , Miguel 1963 Aztec Thought and Culture . Norman : University of OklaLoma Press . 541 Levi-Strauss, Claude 1945 Structural Analysis i n Lir1guis t ics and Anthropology . In Structural Anthropology . New York : Basic Books . pp . 31-54. Lewis , Oscar 1942 The Effects of Whi t e Contact Upon Blackfoot Culture With Special Reference to the Role of the Fur Trade . Monographs of the Alilerican Ethnological Society, No . 6 . Seattle : University of 1.1/asbington Pre ss . Lore nz , Konrad 1966 On Aggression . New York : Harcourt , Brace , and World. Louch , Alfred R. 1966 Explanation and Human Action . Berkeley : University of California Press . Lundberg , Ferdinand 1968 The Rich and the Super- Rich . New York: Lyle Stuart. McEwen , William J . 1963 Forms and Problems of Validation i n Social Anthro- pology . Current Anthropology 4(2 ):1 55-1 83. Malinowski, Bronislaw 1941 An Anthropological Analysis of War. American Jour- nal of Sociology 46 : 52 1-550 . Manners , Robert and David Kaplan 1968 Theory and Method i n Anthropology . Chicago : Aldine. Maybury- Lewis , David 1967 The Savage and the I nno cent . London . Meehl, Paul E . 196 7 Theory- Testing i n Psychology and Physics : a Me thodo- logical Paradox . Philosophy of Science 34 : 103-115 . Montagu , M. F . Ashley 1965 The Human Revolution . New York : Bantam Books . 1968 Man and Aggression. New York : Oxford Uni versity Press . Morris, Desmond 1969 The Human Zoo . New York : Dell. Mueller, John H. and Karl F . Schuessler 1961 Statistical Reasoning in Sociology . Boston : Hov_ghton Miffl ·_n . 542 Murdock , Gorge Peter 1949 Social Structure . New York : Macmillan . 1957 World Ethnographic Sample . American Anthropologi s t 59 : 664-687 . 1966 Cross- Cultura l Sampling . Ethno logy 5: 97- 114 . 1967 Ethnographic At l as . University of Pittsburgh Press . 1968 World Sampling Provinces . Ethnology 7 : 305- 326 . Murdock , George P . and Di ana 0 . Morrow 1970 Subsistence Economy and Supportive Practices : Cross - Cultural Codes 1. Et~1ology 9 : 302- 330 . Murdock , George P . anti Douglas R. White 1969 Standard Cross-Cv.ltural Sample . Ethnology 8 : 329- 369 . Murdock, George P . and Suzanna F . Wilson 1972 Settlement Patterns and Community Organizat ion : Cross- Cultural Codes 3 . Ethnology 11 : 254- 295 . Mur_phy , Robert 1957 I ntergroup Hostili ty and Social Cohesion . American Anthropol ogist 59 :1 018-1035 . Hurphy , Robert F . and Buell Quaj_n 1955 The Truma i Indians of Central Brazil . Ameri can Ethnological Soc i ety Monograph 24 . Myrdal , Gunnar 1971 Ec onomic Theory and Unde rdeveloped Regions . New York : Harper Torchbooks , Harper and Row . Nammour , J ami l 1 973 Resemblances and Universals . Mind 82 (328 ):51 6- 524 . Narn.mour , Valerie Wheeler n . d . Women i n War . Unpub . Ms . Naroll , Raoul 1 96 1 Two Solutions to Galton 1 s Problem . PhilosophJ' of Science 28 : 15- 39 . 1964- A Fifth Solution to Ga lton 1 s Problem. American Anthropologist 66 :863-867 . 1964 On Ethni c Unit Classification . Current Anthropology 5: 283-312 . 1966 Does Military Deterrence Deter? Trans- a ction 3(2) : 14-20 . 1969 Deterrence in History . I n Theory and Research on the Causef-1 of War. Dean G. Pruitt and Ri chard C. Snyder , Eels . Englewood Cb ffs : Prent ice - Hall . _pp . 150-1 64- . 543 1970 A Standard Ethnographic Sample : Preliminary E ition . Current Anthropology 11(2) : 235-248 . Naroll , Raoul and Roy D 1 Andrade 1963 Two Further Solut ions to Galton 1 s Problem . Ameri can Anthropologist 65 : 1053- 1067 Newcomb , W. W. , Jr . 1950 A Re - Examination of the Cause of Plains Warfare . American Anthropologist 52 : 317- 330 . 1960 Toward an Unders tanding of War . I n Essays in the Science of Culture in Honor of Leslie White . Gertrude Dole and Robe rt Carne iro, Eds . New York : Thomas Y. Crowell . pp . 317- 336 . Nie , Norman , Dale H. Bent , and C. Hadlai Hull. 1972 Stat istical Package For the Social Sciences . New York : McGraw- Hill . Ni elsen , H. A,, 1967 Met hods of Natural Science . Englewood Cli ffs : Prentice- Hall . Ot t erbein, Keith F . 1968 I nternal War : A Cross-Cultural Study . American. Anthropologist 70 : 277-289 . 1970 The Evol ution of War . Pittsburgh: HRAF Press . Otterbein , Keith F . and Charlott e Swanson Otterbein 1965 An Eye For an Eye , A Tooth Fo r a Tooth : A Cross- Cultural Study of Feuding . American Anthropologist 67 : 1470-1 482 . Pelto, Pertti 1970 Anthropological Research : the Structure of Inquiry . New York : Harper a nd Row . Popper , Karl R. 196 5 Conjectures and Refutations : the Growth of Scien- tific Knowledge . New York : Harper Torchbooks , Harper and Row. Prasad , Devi 1973 Public address at Calif ornia State University , Sacramento, Apr i l. Rappaport , Roy A. 1968 Pigs For the Ancestors . New Haven : Yale University Pre ss . 544 Reitsma , H. J . 1972 Letter to the Editor . Nat ural History 81 (5 ) :4- 6 . Rohner , Ro nal d P . and Pertt i J . Pelto 1970 Sampling Methods : Chaney and Rui z Revilla , Com- ment 2 . American Anthropo logi st 72 : 14 53-1 456 . Rorwi k , D. M. with L. B. Shettles 1970 Your Baby ' s Sex : Now You Can Choose . New York : Dodd Mead . Russell , Elbert W. 1972 Factors of Human Aggress ion : A Cros s - Cultural Factor Analys i s of Characterist ics Re l ated to Warfare and Crime . Behavior Science Notes 7(4) : 275- 312 . Sahlins , Marshal l 19 6 1 The Segmentary Lineage : An Organizat ion of Preciatory Expansion . American Anthropol ogist 63 : 322- 345 . 1968 Tribesmen. Englewood Cliffs : Prentice- Hall. Service . Elman R. 1960, The Law of Evo l utionary Pote ntia l . I n Evolution and Cultuxe . Marshall D. Sahl i ns and E . R. Service , Eds . Ann Arbor : Uni versi ty of Mi chigan Press . pp . 93-1 22 . Shatt le s , L . B. 1970 Factors I nfluencing Sex Ratios . I nternat i onal J our- nal of Gynaecology and Obstetrics 8 : 643-647 . Sipes, Richard G. 1973 War , Sport s , and Aggre ssion : An Empiri ca l Te s t of Two Rival Theories. American Anthropologist 75 : 64- 86 . Slater , Philip 1970 Tbe Pursuit of Loneliness : American Culture at the Breaking Point . Boston : Beacon Press . Special Study Group 1967 Report From Iron Mountai n On the Po ssibility and De sirabi.li ty of Peace . New ,York : Dial Press . Spradley , James and Davi d McCurdy 1971 Confoi~mi ty and Conflict . New York : Litt le Brown . S1unner , William Grahai:1 1 911 War and Ot her Essays . New Have n : Yale Uni versity Press . Also in War . Leon Bramson and G. W. Goethals , Eds-. - New York : Basic Books , 1968 . pp . 205- 228 . 545 Textor , Robert B. 1967 A Cro ss- Cultural Swnmary . New Have n : HRAF Press . Tiger , Lione l 1969 Men I n Gr oups . New York : Random House . Tinberge n , N. 1968 On War and reace I n Animal s and Man . Science 160 Tuden , Arthur and Catherine Marshall 1972 Political Organi zation : Cross-Cultural Codes 4 . Ethnology 11 : 436- 464 . Turn ull, Colin 1973 The Mountain People . London : Cape . Turney-High , H. H. 1949 Primitive War , I ts Pract ice and Concepts . Columbia : Univers i ty of South Carolina Press . Vayda , Andrew P . 1960 Maori Warfare . Polynesian Society Maori Monographs no . 2 . Wellington , New Zealand : The Polynesian Society . 1 96 1 Expansio n a nd Warfare Among Swidden Agricultural- i sts . American Anthropologist 63 : 346- 358 . 1968 Hypotheses About Functions of War . I n Wa r: the Anthropology of Armed Conflict and Aggression . M. Fried , M. Harri s , and R. Murphy , Eds . Garden City , New York : Natural His t ory Press . 1969 The Study of the Causes of War , With Specia l Refer- ence to Head-Hunting Ra i ds i n Borneo . Eth..nohistory 16(3 ) : 21 1-224 . Walter , E . V. 1969 Terro r a nd Resis t ance . New Yor~ : Oxford University Press . ,fa ltz , Kenneth N. 1959 Man , t he State and War . New York : Columbi a Univer- sity Preas . Warner , W. L. 1940 Murngi n Warfare . Oceania ·1: 457- 4 77 . Washburn , S . L. anQ D. A. Hamburg 1969 Agg1·ess i ve Behavior i n Old World Monkeys and Apes . I n Prima te s . Phyllis C. J ay , Ed . New York : Holt, Rj_nehart and Wi ns ton . pp . 458- 478 . 546 I/ells , Donald A. 1967 The War Myth . Ne w Yor·k : Pegasus . Whi e , Leslie A. 1949 The Science of Culture . New York : Grove Pres s . 1959 The Ev olution of Culture . New York : McGraw- Hill . Whiting , John W. M. arid I rving L . Chi ld 1953 Ch i l d Tra i ning a nd Persona lity . New Haven : Ya le Univers i ty Press . Winch , Peter 1958 The I dea of a Social Scienc e . London : Routledge and Kegan Paul . 1964 Understanding a Primitive So ciety . American Philo- sophi cal Quarte r ly 1 : 307- 324 . Wright , Quincy 1965 A Study of War . Second edition . Uni versity of Ch icago Press . 5L Carolyn C. Larsen