rb lee 1972 kung spatial orgn ecol & historical perspective

8/9/2019 RB Lee 1972 Kung Spatial Orgn Ecol & Historical Perspective

1/24

!kung Spatial Organization: An Ecological and Historical PerspectiveAuthor(s): Richard B. LeeSource: Human Ecology, Vol. 1, No. 2 (Sep., 1972), pp. 125-147Published by: SpringerStable URL: http://www.jstor.org/stable/4602251

Accessed: 17/07/2010 07:15

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

http://www.jstor.org/action/showPublisher?publisherCode=springer.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

Springeris collaborating with JSTOR to digitize, preserve and extend access toHuman Ecology.

http://www.jstor.org
http://www.jstor.org/stable/4602251?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/action/showPublisher?publisherCode=springerhttp://www.jstor.org/action/showPublisher?publisherCode=springerhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/4602251?origin=JSTOR-pdf


2/24

Human Ecology, Vol. 1, No. 2, 19 72

!Kung SpatialOrganization:An Ecologicaland HistoricalPerspective'

RichardB. Lee2ReceivedMarch15, 1972;revisedMay31, 1972

The ecological and social basesof spatialorganizationamonghuntersand gatherersareexamined.After criticizingthe patrilocalband model of social organization, he authordocuments the flexible, nonterritorial roupingsof the !KungBushmenof Botswanaandrelatesthem to rainfalland surfacewaterscarcityand variability.The paper goes on toconsider the effects of extra-Bushmancontacts on the breakdown of sociospatialorganizationand finds that the observedflexibility occurredin both the pre- and thepostcontactperiods. The inal sectionattemptsto relatetheanalysis o general ssues. Threeareas that need furtherworkif a more validmodelof hunterspatialorganizations to bedevelopedare theproblemsof timeperspectiven research, daptation o long-term limaticvariability, ndcritical hresholds f populationdensity.

INTRODUCTIONThis is a contribution to the study of how huntersandgatherersorganizethemselves n space and how this organizationadaptsto variations n populationand resources.Using fields data from the contemporary !KungBushman ofBotswana, I examine the nature of the associationbetween social groupsandtheir space, and I searchfor the ecologicalandsociologicaldeterminantsof thisassociation.Since there has been extended controversy in anthropology on thequestion of sociospatialorganizationof hunter-gatherers,t is important at theoutset to define the problemto be solved and to pinpoint possiblesources of

Supported by the National Institute of Mental Health (U.S.), the National ScienceFoundation (U.S.), and the Wenner-Gren Foundation for Anthropological Research.1See the Appendix for a brief description of the pronunciation of the Bushman languages.2University of Toronto, Toronto, Ontario, Canada.

125?) 1972 Plenum Publishing Corporation, 227 West 17th Street, New York, N.Y. 1001 1.


3/24

126 Leeconfusion. All hunting and gatheringpeoples live in organizedgroups that movefrequently through their ranges. Most modern hunter living groups aresmall-under 200 people-and these groupsare observed o move their campsitesfrom two to ten times per year. Theexistence of a group and a space necessarilyimplies the existence of two kinds of boundaries: social and spatial. A socialboundarycan be measuredaccording o how open or closedthe groupis. At theopen extreme, individuals move at randomwithin a space, encountering oneanotherfor brief periods,then movingon. At the other extreme, thereis a seriesof tightly organized groupswhose members stay together throughout the yearwith minimal interchange with other groups. A spatial boundary can bemeasured along the dimension of overlap/nonoverlap.Imagine a large spacecontainingfive groups. At the "overlapping" xtreme, all five groups have freeaccess to the entire space; at the nonoverlapping xtreme, the five groups dividethe space into five exclusive sectors. In an intermediate condition, each of thefive groups has a core area which is theirsalone, while they share the rest of thespace with their neighbors. These two kinds of boundaries are illustratedinTable I (c.f. Yellen and Harpending,1972).The distinction between social and spatial boundaries s a necessary one.Open groups may have nonoverlapping territories but still accommodatemovement of personnel across the boundaries,and it is at least theoreticallypossible for a closed group to share overlappingranges with their neighbors.Much confusion has arisen from the fact that group boundaries and landboundaries have not been kept separate in analyses of hunter-gathererorganization. As we shall see, both kinds of boundaries are fluid forcontemporaryhunter-gatherers.

A second source of confusionhas been the failure to distinguishbetweenthe behaviorof groups in their space and the conceptions or folk view of thepeople about themselves and their land. The latter type of data, thoughimportant, is at best an imperfect reflection of the actual arrangementsofpersons on the ground. In this paper,my prime concern is with the behaviorof

Table I. Dimensions Along Which Social and Spatial Boundaries MayVary in Hunter-Gatherer Groups

open closed(random movement) (no interchange)Social boundary K- 9

overlapping nonoverlapping(shared) (exclusive)Spatial boundary


4/24

!KungSpatialOrganization: n Ecologicaland HistoricalPerspective 127groupsand not with their ideology, although aterin the study someideologicalaspectsareintroduced for example,p. 142).Havingdefined the dimensionsalong whichsocialandspatialorganizationmay vary, we must now considerwhat formationsare actuallyobserved n theethnographiesof hunter-gatherer ocieties. Ethnographically,here is no caseknown of a society in which the members move randomly in a totallyoverlappingspace. Toward the opposite extreme, however, there is apparentevidence in a number of societies for tightly organizedgroups maintainingexclusive territories. Such groups have been reportedfrom severalpartsof theworld and have been lumped under the generalrubricof the "patrilocalbandmodel" of hunter-gathererocialorganization.In its essentials, he patrilocalbandis basedon threeorganizingprinciples:(1) band exogamy (everyone must marry someone from outside the group),(2) patrilocalpost-marital esidence women move into other groupsat marriage;men remain together and bringtheirwives in), and (3) band territoriality eachgroup controls a space, moves within it, and defends it against outsiders)(Radcliffe-Brown,1930; Steward,1936, 1955; Service,1962, 1966).The operation of these principles results in a situation not unlike the"closed group/nonoverlapping pace"extreme of our model. This is not to saythat no interchangeoccurs: womenmovebetweengroupsat marriage, ndall ofthesesocietiesareacknowledged o haveformalizedvisitingarrangementsor thepurposeof carryingout ritualsandtrade.Nevertheless, he dominant mpressionone gets from accounts of patrilocalbands s one of semi-isolated,male-centeredgroups,encapsulatedwithinterritories.This patrilocalterritorialexogamousbandor "horde,"as Radcliffe-Browncalled it, has proven to be a remarkablypersistentconstruct in the study ofhunting and gathering peoples. For Radcliffe-Brownin 1930 it was "theimportantlocal groupthroughoutAustralia" 1930: 35); othershave endorsedthis view quite recently and have presentedsome ethnographicdatain supportof it (Stanner, 1965; Birdsell, 1970). Still others have sought to establish thepatrilocal form as the basic grouping not only for Australiansbut for allhunter-gatherersresentandpast (Service,1962, 1966; Williams,1968).Whydo so many analystscontinue to be attracted to the patrilocalbandmodel? At least part of its appeal is that it is an elegant and parsimoniousformulation. The society is seen as beingstructuredby the interactionof a smallnumber of jural rules: territorial ownership by males, band exogamy, andviripatrilocalpostmarital residence. Similarly, the spatial arrangementsareextremely neat: a mosaic of territoriesarrangedn a honeycomb pattem, eachcontainingits land-owninggroup. This apparentneatness and parsimonyhaveproved especially useful to model-buildersseeking to characterizehunter-gatherergroup structurein an economic way for computerand mathematicalsimulation.


5/24

128 Lee

However, the problemis that a society based on these ruleswould finditself in severe adaptive difficulties. The patrilocalband makesscant provisionfor a number of basic featuresof hunter-gathererife, such as the necessity toeven out demographicvariationin sex ratios and family size, the continuingnecessity to adjust group size to resources, and the desirability of resolvingconflict by fission.In contrastto the recent advocatesof the patrilocalmodel, I argue n thepresentpaperthat this formof organization s empiricallyrareandthat its raritycan be made intelligibleby a careful ecological analysis.The model of !KungBushmendescribed n the paperappearsto account well for the observedfactsand, moreover,can be appliedto the analysisof other hunter-gathererases.Asimilarcritique of the patrilocalmodel based on !KungBushman materials spresentedby Yellen andHarpending1972).We are of coursenot the first to questionthe universalityof the patrilocalband amonghunter-gatherers.This was done as early as 1936 by Steward 1936,1938, 1955), and criticismsof the patrilocalmodelhavebecome morefrequentin recent years (e.g., Helm, 1965; Fiatt, 1962, 1966, 1968; Leacock, 1969;Meggitt, 1962: 70-71; Woodburn,1968; and variousauthors in Damas, 1969,and Lee and Devore, 1968). According to these criticisms,a relativelyopen,social groupwith overlapping haredterritoriesseems to be the prevailingormamongcontemporaryhunter-gatherers.Yet the flexiblegroupalternative o the patrilocalmodel is not without itsdifficulties. Though different, it appearsat first glanceto be equally unattrac-tive. The recent observations of hunter-gatherergroup structure present anapparently chaotic picture, in which a person may live wherever and withwhomeverhe or she pleases. Thesituationis furthercomplicatedby the fact thatall living hunter-gatherer roupshave been strongly affected by contact withoutside peoples. Service(1962) goes so far as to arguethat the flexibility ofmodernhunter groupings s strictly an artifact of acculturationandbreakdown.In short, we are stuck with a complex set of messy facts collected inambiguousfield situations,and the job of the analyst is to makesenseof them.Since so much of the recent controversy has focused on the AustralianAboriginal material, it may be refreshing to take a different but equallyinterestingcase-that of the !KungBushmanhunter-gatherersf southernAfrica.

!KUNGLOCALGROUPSAND THEIRSPACEWewill considerthe Dobe Area !Kungof Botswana including he /du/da!Kung)and the closely relatedNyae Nyae !Kungacross the borderin Namibia(illegallyoccupiedby South Africa).The latter people have been described n abrilliant series of ethnographiesby LornaMarshall,n the films and writingsofJohn Marshall,and in the writingsof ElizabethMarshallThomas.Aspectsof theDobe Area !Kung have been investigated by a number of authors. The


6/24

!Kung Spatial Organization: An Ecological and Historical Perspective 129description to follow is drawn from the Dobe material, although the mainfeatures apply to the Nyae Nyae !Kung as well. Where specific points ofdisagreement xist, they are explicitly stated.

What are the majorfeatures of !Kunglocal groupings, and ownership,group affiliation, and boundarymaintenance?The basic local groupings a camp(Marshall's and), whichis a noncorporate,bilaterallyorganizedgroupof peoplewho live in a singlesettlement and who move togetherfor at least part of theyear. At the core of each of these campsaretwo, three,or moresiblingsand/orcousins, both male and female, who are generally acknowledgedto be theowners (K"ausi) of the waterhole. Aroundeach waterholeis a bloc of land orn-!ore. This contains food resources and other water points and is the basicsubsistenceareafor the residentgroup. LornaMarshall 1960: 344 ff)has arguedthat the ownership of eachwaterhole resides n the personof a bandheadmanwho is always male and who inheritshis position patrilineally,but my ownresearch ndicates that no headmenexisted amongeither the Dobe or the NyaeNyae !Kung Bushmen.Instead the sibling-cousing roupof K"ausicollectivelyheld the waterhole. (Documentationon the problem of the headman will bepresentedelsewhere.

Specific groups have histories of association with their waterholes thatvary from a few years up to several decades in duration. Rarely does thisassociation go back as far as the grandparentgenerationof the oldest peopleliving. To put it another way, a first approximation of the "half-life"of agroup's enure at a waterholecanbe estimated at 30 to 50 years.An individualmay inherit his or her n!ore from his father'sor mother'sfamily;or he or she may inheritit fromboth parents,or fromneither.There s adiscerniblepatrilineal endency, at least amongmales. A surveyof 151 men overthe age of 15, representing90%of the adultmale residents of the Dobe area,gave the breakdownon inheritanceof n!ore shown in Table II. We see that whilethe majority take n!oresnot in a strictpatrilinealway, inheritance rom father sthe most frequent single alternative.Unfortunately, comparabledata were notgathered or females.

Tsble II. Patrilineal vs. Matrilineal Inheritance of Localityas Reported by 151 !Kung MenInherited n!ore from n SFather 60 39.7Mother 40 26.5Both father andmother 16 10.6Neither parent 21 13.9[Doesn't know, or no n!ore] 14 9.3Total 151 100.0


7/24

130 Lee

The composition of groups at any one time shows little or no patrilocalemphasis.In fact, because of the frequencyof bride service,thereis even a slighttendency toward matrilocality(cf. Leacock, 1955). Of 114 currently marriedcouples, 22 were living with husband's parent(s); 24 were living with wife'sparent(s),while 12 werelivingwith both husband'sand wife's parents.A further15 couples were living neolocally, while in the case of 41 couples, neitherhusband'snor wife's parentswere alive. Those couples in whichboth had livingparents were observed to pay frequent visits to both sides. This and otherprocesses of intergroup visiting created a fluid situation in which thecompositionof groups changedfrom week to week with the comings and goingsof people (Lee, 1972a).Within a local group's area, subsistence resources are not exclusivelyreserved for the K"ausi and their families. By observing elementary goodmanners,anyone who has a relativein a camp may enjoy the resourcesof thearea around the camp. Within the camp, food is sharedin such a way'thateveryone-residents and visitors alike-receives an equitable share (cf L.Marshall,1960, 1961).In moving through the annual round, the !Kung groupingssatisfy theirsubsistence requirements with surprisingly little friction with neighboringgroups. Subsistence space is bounded, but these boundariesare vague and notdefended.In fact, a frequentpatternis for groupsfrom two or more waterholesto join forces for the joint exploitationof a major resource,such as tsin beansormongongo nuts. And duringthe winter dry season, it is common to find fromtwo to six differentgroupscampingtogetherat a permanentwaterhole.

ENVIRONMENTAL AUSESOF FLEXIBILITYHow do we account for this flexibility of group structure? Before weconclude that it is due to the effects of recent contact and acculturation,we

shouldconsider what effect pre-existing eaturesof the environmentwouldhaveon groupstructure.Two permanent eaturesof the Kalahari nvironmentappearto be salient: first, the high variability in rainfall and, second, the sparsedistributionof standingwater in the northernKalahari.Let us consider each ofthese features n turn.

RainfallVariabilityAverage annual rainfall has little meaning in an environmentin whichrainfallmay vary by as much as 300%from year to year. Duringthe 1963 to

1964 rainy season, we recorded 239 mm (9.4 inches), while during 1967 to1968, 597 mm (23.5 inches) was recorded. Giventhis variability, t would bemore useful to discussrainfall n terms of extremes-for example, tabulating henumber of years out of every ten in which droughtis experienced.Maun,300


8/24

!Kung Spatial Organization: An Ecological and Historical Perspective 131km by air southeast of Dobe, is the nearest weather station for which long-termrainfallrecords are available.Figure 1 illustratesthe fluctuations n rainfall hatMaun has experienced over the last half century, Table III summarizes hesedata. In a run of 46 years, droughtoccurred n 17 years(37%o),nd,of these, 12

, . g 10 9~~~~~~~~~~~~ROUGHT

900-_

800

0

1922-23 1930-31 1940-41 1930-sI 1960-61 667-66

RAINFALL YEARFig. 1. Variations in annual rainfall at Maun, Botswana, July 1922 to June 1968 (mean

462.3 mm based on 46 years of records).Table III. Drought Frequencies and Intensities at Maun Botswanaover a 46-Year Period (Lat. S20?00' Long. E23?26')

nYears of normal rainfallP7 29 63Year of droughtb 17 371st degree (mild) 5 112nd degree (severe) 11 243rd degree (very severe) 1 2Total 46 100'7Normal rainfall is annual rainfall more than 85% of mean.bDefinitions of drought severity adopted from Wellington (1964:

40-43): 1st degree, 70 to 84% of mean rainfall; 2nd degree, 55 to69% of mean rainfall; 3rd degree, less than 55% of mean rainfall.


9/24

132 Leeyears (26%) were classified as severe drought when less than 70% of averagerainfall occurred. In other words, the probabilityof droughtoccurringat Maunis about 2 years in 5, andof severedrought,1 year in 4.The situation for Dobe would be even more acute since rainfall s loweroverall (about 350 mm at Dobe vs. 462 mm for Maun) and therefore moreerratic in annual fluctuation.Judging from our experiencesovera 7-year period(1963 to 1969) and that of the Marshalls 1952 to 1959), drought conditionsprobably characterizeabout half the years. For reasons whichare not clear, thenorthern Kalahariappears to experience alternatingrunsof good years and ofbadyearsof varying ength.A second source of variabilityis the differencein rainfallfrom place toplace in a single month or season.In a clusterof five stationswithin an area 200km across, annual totals may be comparable,but the fall in a given month mayvary from place to place by as much as a factor of 10. Table IV shows therainfall for 1966-67 rainy season at five stations in the Ghanzidistrict 300 kmsouth of Dobe. In the early rainsof October to December,this local variation scrucial, since it is these rainswhich largely determine the overallsize of the wildfood harvest later in the season. For example, in November, Kalkfonteinreceived only 3.5 mm while Scarborough,50 km away, received34.0 mm. As aresult, the desert may be bloomingin one area while a few hours'walk away itwill still be parched.It is to such variableconditions that the spatial organizationof the !KungBushmen must adapt. Theirs is a long-termsolution to the problem,and it is

Table IV. Local Variations in Rainfall Among Five Localities in the GhanziDistrict, Botswana, 1966-67 Rainy SeasonRainfall in mm at

Kalkfontein Scarborough Ghanzi Oakdene Cume1966July - -August - - 6.3 - 0.5September - 12.0 41.2 14.5 25.7October 16.0 10.5 24.1 - 24.5November 3.5 34.0 35.2 43.2 29.0December 73.5 116.0 111.6 90.5 75.41967January 127.0 167.6 242.0 183.0 223.2February 81.0 136.0 139.2 161.5 131.3March 38.0 35.5 14.5 6.3 50.0April 25.0 89.5 6.2 55.4 86.3May 18.5 13.5 9.1 21.0 6.5June - - - - -Total 382.5 614.6 629.4 575.4 652.4


10/24

!Kung Spatial Organization: An Ecological and Historical Perspective 133

unfortunate that an ethnographerwill observe only a small segment of thepattern n any givenyear of fieldwork.WaterSourceScarcity

The sparse distributionof water on the groundis the spatialcorrelateofthe temporal variabilityjust discussed.Because of the porosity of the sandysoils, the high rates of evaporation, and the infrequency of exposures ofwater-bearingrock, standing water points are few and far between. Figure 2shows an areasouth of the AhaHillsandnorth of the EisebValley.This includesthe southernhalf of the Dobe areaand partsof the Nyae Nyae and/du/da areason the west and south, respectively.The areastraddlesthe internationalborderand measuresabout 80 by 80 km for a total areaof 6400 km2 (2500 squaremiles). This entire area contains only five permanent water points, that is,waterholes that hold water throughout the year, and of these three have beenknownto failwithinlivingmemory.

17tA H4,A, lii'NILSn!umdi'~

hxors n.woma

/widum /dwla/Ihab

/ose

DUE ...D U E *" codanaha

z m -*_

/gom - //gona //gum//genl !Al

~~x ,j


11/24

134 Lee

In fact, the area containsa hierarchyof water sources arranged n orderoftheir duration and reliability: I, two have never given out in living memory(/ai/ai, /gam); II, three have not failed in the last fiveyears (/wihaba,#o//gana,/du/da); III, five (at least) arestrongsummerwaters whichmay last throughthewinter of good years (/dwia, n#wama, hxore, !widum, //gum//geni); IV, about50 are seasonalwaterpointsholdingwater from a few days to severalmonths; V,about 100 are mongongo,baobab, and terminalia rees with small reservoirsntheir hollow boles; VI, about 150 are sites in which the water-bearing oot !xwa(Fockea monroi)is found.In general, the fartherdown the hierarchy of water sources they go, theharder it is for the !Kung to satisfy their moisture requirements.Tree water(class V) is usually difficultof accessandoften must be soaked up in a makeshiftsponge or sucked out in a reed straw. (Sip-wells are known by the !KungBushmen but rarely utilized.) Water root (class VI) is evenmore difficult to getat, since the root itself must be dugout from depthsof up to 40 cm (15 inches)and contents of as many as 20 roots must be consumedper day to providetheneeds of one person (cf Lee, 1965: ch. 8). On the other hand, root and treewaters are widely distributedin close proximity to valuablefood sourcesandthereforeare often utilizeddespite their difficulty of access.The !Kung ability to operate successfully in this environment thereforeinvolves them in sorting out a complex set of variables about the currentlocations of food and water, the ease or difficulty of getting it, and thewhereabouts and current activities of adjacent groups. And their subsistenceplans must be continually revised in light of the unfolding rainfall situationthrough the growingseason andbeyond.

PATTERNSOF !KUNGSPATIALORGANIZATIONGiven the ecological conditions, what land use pattern has emerged?

Today (1963 to 1969), internationalpolitics and pastoral occupation haverestructured he land use in non hunter-gathererways. The currentsituationandhow it evolved are interestingquestions in their own rightand will be discussedbelow. In order to understand he hunting andgatheringpattern,we have to doa reconstructionto get a picture of the distribution of landholdinggroupsasthey wereduringthe 1920s, before the Bantu settlement.In Fig. 3 are plotted the landholdinggroupsof the /gam-/ai/aiareasof 40years ago. The number is located on the map in the vicinity of the largestwaterhole each group's area contains. An examination of Fig. 3 makes itapparent that there were many more groups holding land than there werepermanentwaterholes to supportthem. Elevengroups regularlywinteredin thearea,with occasional visits from at least four other groups;yet there wereonlyfive permanentwaterholes and of these only two were really reliable.Elevenis


12/24


n/um , i '-.

? heaxore n#wama

DU iIwidum , /dJ,aodur; hdb

i /os- c h -

DUE m o * ~~~codanaha

/gam *Aon /gum//geril N,i g _ j;S/AlxoaI o/on , , - - -

LEGEND[ 1I-_ -- ----Oon/o _- u~ _ _n>,^ i Permanent Waterholes

---- -- "'/ t GSemi-permonent\_ _ v,Waterholes*Major SeasonalE :g ~~~~WaoteholIes

I!] '*scale o? o kmto n//o!au and dumn/a ,-, _'

Fig. 3. A reconstruction of the distribution of landholdinggroups, 1920 to 1930.

the minimumnumber that operated in this area. There may have been moregroups,especially in /gam. The members of these groupshave moved entirelyout of the area to the larger"magnets"of the Ghanziand Gobabiswhite farms,or to the Tswana and Herero cattle posts around Lake Ngami. Futher, thispresentationdoes not take into account groups that have movedinto the studyarea since 1930, such as the groupfrom/gausa(355) who movedto /ai/ai in the1940s, and groupsfrom the Ghanzi farmswho weremoving nto the /du/da areain the late 1960s (625 and 729).Such a large number of groups could be supported only if there werewidespreadagreementto maintain regularand free access to permanentwater.The orderly, rather evenly spaced arrangement of groups in Fig. 3 ischaracteristiconly of the height of the rainy seasonwhen water and food areavailablethroughout the area. But as the summer waters evaporate with the


13/24

136 Leecoming of the winter, the 11 groups would converge to the patternplotted inFig.4 andsummarized n Table V, column 2.Unlesswater was exceptionally strong,groups2 and 4 wouldjoin group 3at /wihaba, and later in the season most or all of the peoplewouldpay a visit togroup 1 at /ai/ai. At the same time, groups 5, 6, and 7 would converge on*o//gana, while groups 8, and 9, and 10 would converge on /gam. Group 11stayed around/du/da, where it wasjoined by one or two groupsfrom n//o!auand dumn!a(see Table IV footnote). Laterin the season, it was customaryforgroupsat /du/da and#o//gana to pay visitsto relativesat /gam.In good years, groups had the option of wintering n any one of severalplaces: in the home area, at a permanentwaterhole, or visiting relativesatwaterholesoutside the area.Also, the membersof the groupsthat wereprimarilyassociated with one of the two very reliablewaterholes-group 1 at /ai/ai andgroup 8 at /gam-could spend most of their year enjoying the seasonal food

I l-

',o hxortAn, gj -L Sw m a --n/umdi

0hore nVwama

'widum |de i /wlhaba

/ose _DUE

Z's codonoha -

/gam o//ga,ono //gum//genl aI '~~~~~~~~~~~Al

NJ \ 9 I-;EGENDPermanent WoterholesOona

___ r 0 ./, Semi- permanent/du /d a ~ - -~ -+ 5OWaterholesud/do Major SeasonalWaterholes

L, -'scale O kmto n//a/au and dunam

Fig. 4. Patterns of convergence of landholding groups during thedry season and in drought years, 1920 to 1930.


14/24


Table V. Reconstructed Groups of the /ai/ai and /gam Areas, Circa 1920 to 1930aLocation

Group Extreme dry CurrentNo. Informant Summer Winter recourse status(1) (2) (3)

1 396 n!umdi /ai/ai /ai/ai !xumn!a's groupat /ai/ai2 346 /wihaba /wihaba /ai/ai No longer function-and /ai/ai ing; descendentsat /ai/ai3 414 /dwia /wihaba /ai/ai /i!ay's groupand /ai/ai at /ai/ai4 429 !ai /wihaba /ai/ai No longer function-and /ai/ai ing; descendentsat /ai/ai

5 335 6odanaha #o//gana /gam or Still utilize n!ore/ai/ai part of the year;winter at /ai/ai6 543 //gum//geni 4//gana /gam or A few families still/ai/ai utilize n!ore; winter at/ai/ai or /du/da7 363 n. of 4o//gana /gam or No longer function-#0//gana /ai/ai ing; descendentsat /ai/ai8 336 Around /gam /gam No longer function-/gam ing; descendentsat /ai/ai9 580 due or /gam /gam A functioning group#0n! a now based at /du/da10 636? due /gam /gam Moved to South AfricanGovernment settle-ment scheme atTsum!we11 684 /du/da /du/da /gam A functioning groupbased at /du/da

aIn addition, there were several groups who moved into this area from the west andsouth during the summer and may have opted to winter at /gam and/or /ai/ai duringextremely dry years such as Marshall's bands 1 and 2 from /gausa, band 15 fromdum!na, and band 17 from n//o!au (L. Marshall, 1960).

resourcesof their neighbors.Reciprocalaccessto resourcesat all timesensuredthatkey resourceswouldbe availableat criticalperiods.In moderate winter dry-seasonconditions, the 11 groups distributedthemselves at five water points. Under very severe conditions, the groups


15/24

138 Lee

underwent another phase of convergence(see Table V. column 3). Four groupsfell back on /ai/ai, four others on /gam, and three others alternatedbetween/gam and /ai/ai. Thusin the most difficult droughtyears there might be as manyas seven groups at a single waterhole. Such a situation was observed by theMarshallswhen seven groupsconvergedon /gam in the severedroughtwinterof1952 (J. Marshall,1957: 36).Traditionally,waterholessuch as /ai/ai and /gam have played importantroles as entrepots in the economic and sociallives of the !Kung.Even before theBantubuilt theircattle posts there, /ai/ai was a tradingcenter where people fromall points of the compass came to visit, dance, and do hxaro trading (andsometimes fight)./I!ay 414 described he traditionalrole of /ai/ai as follows:

/ai/ai has always been a meeting place for people even before the blacks came.People came from the north, from !angwa and !gose, and from the south from/gam (and from the west from /gausa), stsyed here, did hxaro, drank n!o (achoice wild fruit), ate //"xa (mongongo nuts), and then went back.They asked Kan!o [one of the owners] for permission. They also askedorma!gain [Father of 396]./ai/ai was favored because the water was so big. Choma [men's initiationceremony] was danced here, but the main reason [to meet] was hxaro trade.

People came in all seasons of the year-summer, winter, and spring. Butthey particularly came in spring [!gaa, September-October] when the treeswere in flower [before the onset of the rains]. When the summer pans driedout, they ate !xwa [the water-bearing root]. When the !xwa got thin theycame into /ai/ai. This was because some of the n!ores did not have year roundwater.A closely comparableanduse patternhas been observed n the Nyae Nyaearea as recently as 1952 to 1953 by the Marshall xpedition.In a little-readbutextremely informativework ("Ecology of the !KungBushmenof the Kalahari,"Harvard eniorhonors thesis, 1957), John Marshalldiscussedhow the Nyae Nyaegroupswould arrangethemselves with referenceto water first under "normal"and then under drought conditions. Of the latter, he writes:

In very dry years, more of the bands would be concentrated around/Gautscha, Deboragu, and perhaps Khumsa. I do not know for sure whetherKhumsa is a permanent waterhole. I am sure that there is a permanentwaterhole northeast of /Gautscha to which bands 8, 9, and 10 would shrink ifa winter of desiccation was complete. The distribution of the interior bands insuch a winter season that so utterly rejects all juicy things would probably be:Bands 1, 2, 3, at /Gautscha is the highest yielding waterhole in the area,therefore able to support the 85 people of these bands.Bands 4 and 5 might hold out at Deboragu and would be joined by band 6.Deboragu is a weak water. The 29 people of bands 4, 5, and 6 would probablybe able to survive, however. One man told us, speaking with affection forDeboragu, that, though it may look dry, scratch and you will find water.Band 11 might flee to S'amangaigai and so the people would endure. (J.Marshall, 1957: 32-33)


16/24


HUNTER-GATHERERSPATIALORGANIZATIONThe spatial organization of many hunting and gatheringpeoples wassimilarto that of the two !KungBushmencases cited. Forexample,a divisionofthe Eskimo year into a large-groupphase and a small-groupphase was firstformallydescribedby Mauss n a classicpaper 1906), anddocumentedby manyobservers(Boas 1888; Rasmussen, 1931; Spencer, 1959; Balikci, 1964; Dama,1969; the last source is particularlyuseful). In the case of the centralEskimo,

thetimeof maximumconcentrationwasalsoin the winter,but the environmentaldeterminant was the accessibility of good seal hunting rather than theavailabilityof water. For other Eskimo groups, the maximal aggregationwasassociated with a variety of ecological strategies,as summarizedby Damas(1969: 135-138).Among the AustralianAborigines,the flexible land use patternwas for along time obscuredin anthropological tudiesby a confusion of the patrilinealtotemic group with the on-the-ground iving group.The totemic groupindeedcontrolledrealestateexclusivelybut only for occasionalritual purposes,andnotfor day-to-day iving (Hiatt, 1962; Berndt, 1970). For the latter, the groupthathunted, gathered,and lived togetherwas made up of membersof a numberofpatriclansand exhibited a genealogicalcomposition and an annual pattern ofconcentration-dispersionimilar to that of the Bushmen.In arnhemLandandCape York, the significantecological determinantappearedto be the annualflooding of the plain which causedthe people to congregaten largergroupsonthe seacoast (Thomson, 1939: 209) or on higher interior ground (White andPeterson, 1969; Schrire, in press; see also Hiatt, 1965: 24-29). In desertAustralia,the concentration-dispersionatternhas been knownfor many years.Particularattention has been paid to the maximalgrouping n the form of thecorroboree or ceremonial gathering(Spencer and Gillen, 1899: 271 ff). Theecological significanceof this gatheringhas been pointedout by Meggitt 1962:54-55) and Strehlow(1947: 65). Here, as amongthe Bushman, he environmen-tal determinantwas seasonaldifferences n wateravailability.

Examples could be multiplied: concentration-dispersion nd reciprocalaccess to resourceshave been documented for subarcticIndians(Helm, 1965;Leacock, 1955, 1969), GreatBasinIndians Steward,1938, 1955), andPygmies(Turnbull, 1965, 1968); the case of the Northwest Coast is discussedbelow.However,what is centralto all of these cases is a patternof concentrationanddispersion, usually seasonal, and a set of rules and practices for allowingreciprocalaccessto orjoint exploitationof key resources.The worldwide occurrenceof this patternof spatialorganization n vastlydifferentkinds of environments ndicatesthe degreeto which it was basicto thehunting and gatheringadaptation.Several of the adaptive advantagescan bespelled out. In the case of the !Kung Bushmen,we see, first, that reciprocal


17/24

140 Leeaccess to resources allowed a much higher population density than could besupported f it were requiredthat every n!ore contain a permanentwater source(Fig. 4). Thus in the /ai/ai-/gam areas, we find II groups in occupation insteadof two. Second, the pattem contained a mechanism for responding to localimbalance in food resources. It had the capacity to adjust to conditions ofscarcity and also to conditions of exceptional abundance. Third, the patternoffered many social advantages,not the least of which was the separatingout ofindividuals and groups in conflict, thus keeping the threat of violence to aminimum. (Leacock, 1969:14 cites a very similar set of advantages for theflexibility of Montagnais roupings.)By contrast,the patrilocalpatternof spatial organization hat encapsulatesa group of males with their spouses and offspringwithin a territoryis far lessadaptive. Indeed, it would be difficult to visualize how a patrilocalterritorial-organizationcould function in the Bushmancase. I would predict that such asociety could surviveonly to the extent to which its members could sloughofftheir patrilocality and territorialityand approximate he flexible model outlinedabove. (It is curious that Birdsell, one of the foremost exponents of theecological approach n anthropology,should have chosen to espousea model ofhunter social organization that is as ecologically unviableas is the patrilinealband.)In view of these adaptive advantages,t hardlyseemslikely-as Servicehasargued-that this flexible land use patternis strictly a product of acculturationbroughtabout by the breakdownof aboriginalbands. Flexibility appears o beadaptive in both the precontact and the postcontact situation. In fact, we arenow in a position to trace what actuallyhashappenedto change !KungBushmanland use patternsover the last 80 years.

CONTACTAND SPATIALORGANIZATION: 890 TO 1969Starting in the late 1880s and early 1890s, Tswana pastoralists begancoming out to the /ai/ai-/gam areas from their towns in the east for annualhunting and grazing expeditions. At the end of each rainy season, the various

hunting parties, along with severalgroups of Bushmen, would rendezvousforsome weeks of hunting, dancing,and trading.In the trade, the !Kunggave furs,hides, honey, and ostrich eggshell beadwork, while in return they receivedtobacco, clay pots, iron implements,andEuropeangoods. Whenthe tradingwasdone, the oxen were inspannedand the Tswana drove theirwagonsbackto theeast for the winter.During his periodof initialcontact, an annualconcentrationpoint occurred at this encampmentknown as koloi (ox-wagon, or ox-wagoncampin Setswana).Duringthe 1920s, permanentBantu-speakingettlersbegan to move intothe area, bringingherds of livestock and enlargingand deepeningthe waterholes


18/24

!KungSpatialOrganization:An EcologicalandHistoricalPerspective 141at /gam and /ai/ai. A nucleus of semisedentary!Kungbeganto developat thesetwo points in a process that has been observed worldwide among hunter-gatherers around what L. R. Hiatt has aptly called "the magnets" ofattractiveness.Mission and Government Stations constituted the magnets inAustralia,while, in the northern Kalahari,Bantu cattle posts were the magnets(Lee, 1972b).Prior to Bantusettlement, the !Kunghad spent most of the year movingaroundthe n!ores and a few months camped at the permanentwater.Since thearrivalof the Bantu, a reversepatternhas evolved. Today, many !Kungremainmost of the year camped at /ai/ai and spend only a few months of the yearmovingaroundthe n!ores. In fact, the point of majorpopulationconcentrationin recent yearshas usually coincidedwith the Christmas east offeredthe !Kungby theirBantuneighbors Lee, 1969).The effects of contact on spatial organizationare shown in Fig. 5 (andTable V, last column). Acculturation has produced fragmentation and

A H -A I4

,hxor | - n,#wam a

~~~ .~~/widum w a -___DUED~~~~~~ e. d >*naha ->

/gam E//a / /g _g ,_

Son.a t . u_;_ _ + Permanent Waterholesan/a ~~~~~~~~~SemipermanentWaterholes

/du/ a ~~~~~Majoreasonal/d/j 11_ ; Waterholes5 '. j ~ ~ sCaI ?-~ 5 to?

to n//onau and dumn.a

Fig. 5. Current land use patterns of active landholding groups,1963 to 1969.


19/24

142 Leediscontinuous utilization of n!ores. Four groups have ceased to function assubsistenceunits, havingbecome wholly attached to Bantucattle posts (groups2, 4, 7, and 8). One group, 10-along with many others from the Nyae Nyae,outside our study area-has joined the South African governmentsettlementstation in Tsum!we.Four othergroupsmove in and out of /ai/ai on huntingandgathering ripsof varying ength(groups1, 3, 5, and6).Even though these semisettledgroupsspend most of the year at /ai/ai (orTsum!we),each triesto spendat least a month or two in the home n!ore. Unlikethe Australians, he !KungBushmen do not maintaintotemic sites within theirhome localities. Nevertheless, the ties to the n!ore are certainly based onsentimentas well aseconomicexpediency;this emotional content is expressed nthe following quotation from a young woman memberof group 3 now livingat/ai/ai:

[You see us here today but] you know we are not /ai/ai people. Our truen!ore is East at /dwia and every day at this time of year [November] we allscan the eastern horizon for any sign of cloud or rain. We say, to each other,"Has it hit the n!ore?" "Look, did that miss the n!ore?" And we think of therich fields of berries spreading as far as the eye can see and the mongongo nutsdenselv littered on the ground. We think of the meat that will soon be hangingthick from every branch. No, we are not of /ai/ai; /dwia is our earth. Wejustcame here to drink the milk.In only two cases(groups 9 and 11), are the groupsusing theirn!ores inanythinglike the traditionalmanner.And in the last 3 years,even thesegroupshave been affected. South African police patrolshave ordered these groupstoconfine their camps and activities to within a close radiusof the borderso thatthey can be easily checked up on. This has produced two ratherbizarreeffectson spatialand social organization:first, there is a highly unusuallinearpatternof land use as the groupsmove up the borderroadfromcampto campandthendown again,and, second, there are abnormally argegroupsof 90 to 120 people

camping together at times of the year when one would expect them to bedispersedinto much smallergroups.The !Kungsay they are afraid to dispersefor fearthat the police patrols will go out after them(cf Fig. 5).In short, contact has produced in !Kungland use a spectrum of effectsincludingfragmentationand sedentism in some groups and consolidation andmobility in others. The actual changes in land use can be accounted for by acombination of economic and political factors, although common to allsituations is the introduction of an economic "magnet"and along with it anoutside jural authority (Lee, 1972a, 1972b). The highly flexible spatialarrangements of today appear to be a continuation of flexible spatialarrangements f the precontact era. And these flexiblearrangementsn tum areshown to be adaptationsto the perennialproblems of the arid environment:recurrentdroughtandscarcityof surfacewater.


20/24

!Kung Spatial Organization: An Ecological and Historical Perspective 143CONCLUSION

It remainsnow to deal briefly with severalmethodologicalissues.First, Iwant to specify the operationsof the method usedin this paperso that it can beappliedto other ethnographiccases. In the analysisof a givencase,we consider,first, how the environmentvariesspatially n terms of the uneven distributionofresources and, second, how the abundanceand distributionsof these resourcesvary through time. The resources that vary may be water supply, gamepopulations, salmon runs, vegetable foods, or other factors. Each casewill haveits own constellation of factors. Then, invoking Elton's concept of MinimalorEconomic Density(Elton, 1927), we delineate the minimumarea that a groupofpeople has to maintain access to in orderto ensure its survivaln the mediumand long run. For example, a hunter-gatherergroup may be able to satisfysusistence requirementswithin 100 km2 for 4 years out of 5 but it will still goout of the businessunless it has access to a much largerareaduing the fifth year.And in order to ride out environmental luctation over the course of 50, 100, or200 years, the area to whichthe groupmust maintainaccessmust be evenlarger,probably on the order of 10 times the area it covers in a single good year.Maintainingaccess to such a largearea s reallya questionof maintaining ordialworking relations with one's neighborsoccupyingthe space.So the environmen-tal problemhas a social solution.However, ittle of this long-termperspective s visibleto an observer.Whenan observerarriveson the scene andfinds a huntingandgatheringpopulation na state of constantmotion, he may be initially puzzled by this mobility, sincethe people appearto be movingeven more frequently thannecessaryto exploitwhat appearto be rather stable resources. Faced with such a set of facts, theobserver is liable to attribute this mobility-as Service does-to breakdownofaboriginalbandsthrough contact (Service, 1962: 108), or he may conclude withTurnbull that the mobility is socially determined and has nothing to do withenvironmentalactors(Turnbull, 1965: 177-178).Both these interpretations uffer from the shorttime perspectiveenforcedby the limitations of anthropological ieldwork.An ethnographern his stint inthe field observesone or at best two repeatsof the annual round, and on thisbasistries to generalizeabout Pygmy life or Eskimo life. But we have seen thatthere is no such thing as a typical year for a hunter-gatherer opulation.Theiradaptationis a long-term one, and the observer can catch only a very shortsegmentof the whole in a year.Whenwe see hunters moving widely about their range, in the apparentabsenceof ecological necessity, we arewatchingintergroup,economic relationsthat takeyearsandgenerations o unfold. Keepingup distantsocial ties againstapossiblefutureneed and visiting neighborswho owe favorsfrom previousyearsare only two of the factors that set hunter groups in motion. The ostensible


21/24

144 Leepurposes are social, but the underlyingrationale is adaptive.This may alsohelpto explain why hunter-gatherersrade beads in exchange for beads with theirneighbors. The trade item in the perspective is a facilitating device formaintaining relations that may be ecologically crucial over the long run.Similarly, when an investigator reports an environment which is withoutsignificant regional or temporal variation (Turnbull, 1965, 1968), we maysuspect that he has not looked into the mattercarefullyenough or long enough.Population density is also a key variable. An adequate analysis ofenvironmentalvariability must also plot the minimal subsistence areas forvarying levels of population density. A resource area that looks quiteundifferentiatedfor five persons per 100 km2 (13/100 square miles) may behighly differentiated when the population grows to 25 persons per 100 km2(65/100 squaremiles). For example, a population in the process of moving ntoa new area and occupying it at low densities would be initially immune tofluctuations in key resources,and its members thus mightmanagetheir affairswithout elaborate arrangementsfor reciprocal access. But their populationwould grow after several generations to the level where environmentalfluctuation would threaten their survival.Long before this point is reached,however, one would expect that the necessarymechanismsof reciprocalaccesswould have evolved. At this point, other forces tending to limit populationdensity become operative, and these serve to prevent the population fromthreatening he overall evel of resources Lee, 1972b).Finally, it might be argued that the method presented here is applicableonly in the most marginalenvironments with maximum unpredictabilityofresources.AgainI invokepopulationdensity.Since all environmentsvary,for allenvironments there will be a certain thresholdof populationdensity at whichpoint the resource base will become unpredictable.It is extremely interestingthat concentration-dispersion and use and rules for reciprocal access toresourcesare found evenin the "richest"environmentsamongthe most affluentof the world's recent hunter-gatherers.The Indians of the Northwest Coast(Drucker,1955) annually dispersedfrom their largewintervillagesinto smallersummer settlements located nearer to prime fishing sites. And it was on theNorthwest Coast that the pioneer researchin the problem of environmentalvariation in relation to spatial organizationwas carriedout by WayneSuttles.Suttles (1960, 1962, 1968) has shown that even the richenvironmentof coastalBritish Columbiawassubjectto severe ocal and annualvariation n salmon runs.Without the annualdispersionand the reciprocalaccess to resourcesofferedbyintervillagefeasts and potlatches, many villages of coastal Indianswould havegone out of business see also Vayda, 1961; Piddocke, 1965).If the method and the argumentpresented n this paperhavemerit, then itmay be appropriatefor us to discardmodels of prehistoricpopulations thatencapsulateeach group of males within a territory and to consider instead a


22/24

!Kung Spatial Organization: An Ecological and Historical Perspective 145more dynamic model in which interlocking aggregationsof persons undergocontinualreshufflingof groupsin response to short- and long-termenvironmen-tal fluctuations and to changes n population density.

APPENDIXThe Bushman anguagesare characterized y clicks, sounds producedwithan ingressive airstream when the tongue is drawn sharply away from various

points of articulationon the roof of the mouth. The four clicks, along withexamples and some Englishequivalents, ollow:/ Dental click as in /ai/ai, /du/da (in spoken English,this sound denotes amildreproach,written tsk, tsk)0 Alveolarclick as #o//gana, #on!a.

Alveopalatal lick asin IKung, i!ay.// Lateralclick as in #o//gana, //gum//geni (in spoken English, his sound isusedin some dialectsto urgeon ahorse).

Bushmanwordsmay be pronounced by simply dropping he clickin caseswhere a consonant follows the click or by substitutingt or k wherethe click isfollowed by a vowel or w. For example, for /i!ay readtikay, forgam readgam.

ACKNOWLEDGMENTSI would like to thank Sa//gai, /i!ay, #oma !om !gowsi, /ase, //aihan!a,and

K"au!oma for their patience in describing past land use patterns in the/ai/ai-/gam areas, and for many fruitful suggestions.Mark Dornstreich,PatDraper,MichaelHarner,Henry Harpending,JuneHelm, Nancy Howell, PatriciaKoten, Eleanor Leacock, and John Yellen have made helpful comments onearlier versionsof the paper.R. J. Anderssonand the officers of the BotswanaMeteorologicalOffice kindly provided he rawdataon MaunandGhanzirainfall.Lois Johnson drew the figures.The materialfor this paperwas collected duringthree years of field work with the !KungBushmen n 1963 to 1964 and 1967 to1969.

REFERENCESBalikci, A. (1964). Development of basic socio-economic units in two Eskimo communities.Bulletin of the National Museum of Canada 202.Berndt, R. M. (1970). Comment on Birdsell (1970). Current Anthropology 11: 132-133.Birdsell, J. B. (1970). Local group composition among the Australian aborigines: A critique

of the evidence from fieldwork conducted since 1930. Current Anthropology 1l(a):115-142.Boas, F. (1888). The Central Eskimo. Bureau of American Ethnology Annual Report 6:399-699.Damas, D. (1969). Characteristics of Central Eskimo band structure. In Damas, (ed.),


23/24

146 LeeContributions to anthropology: Band societies. National Museum of Canada Bulletin228: 116-138.Drucker, P. (1955). Indians of the Northeast Coast. Natural History Press, Garden City,N.Y.Elton, C. (1927). Animal Ecology. Sidgwick and Jackson, London.Helm, J. (1965). Bilaterality in the socio-territorial organization of the Arctic drainageDene. Ethnology 4: 361-385.Hiatt, L. R. (1962). Local organization among the Australian aborigines, Oceania 32:276-286.Hiatt, L. R. (1965). Kinship and Conflict: A Study of an Aboriginal Community inNorthern Arnhem Land. ANU Press, Canberra.Hiatt, L. R. (1966). The lost horde. Oceania 37: 81-92.

Hiatt, L. R. (1968). Ownership and use of land among the Australian aborigines. In Lee, R.,and DeVore, I. (eds.), Man the Hunter. Aldine, Chicago, pp. 99-102.Leacock, E. (1955). Matrilocality in a simple hunting economy (Montagnais-Naskapi).Southwestern Journal of Anthropology 11: 31-47.Leacock, E. (1969). The Montagnais-Naskapi band. In Damas, D. (ed.), Contributions toanthropology: Band societies. National Museum of Canada Bulletin 228: 1-17.Lee, R. B. (1965). Subsistence ecology of !Kung Bushmen. Unpublished doctoraldissertation in anthropology. University of California, Berkeley, 209 pp.Lee, R. B. (1969). Eating Christmas in the Kalahari. Natural History (December), 14-22,60-63.Lee, R. B. (1972a). The !Kung Bushmen of Botswana. In Bicchieri, M. (ed.), Hunters andGatherers Today. Holt, Rinehart and Winston, New York.Lee, R. B. (1972b). Work effort, group structure and land use in contemporaryhunter-gatherers. In Ucko, P. J., Tringham, R., and Dimbleby, G. W. (eds.), Man,Settlement, and Urbanism. Duckworth, London, pp. 177-185.Lee, R. B. (1972c). The intensification of social life among the !Kung Bushmen. InSpooner, B. (ed.), Population Growth: Anthropological Implications. MIT Press,Cambridge, Mass.Lee, R. B., and Devore, I. (eds.) (1968). Man the Hunter. Aldine, Chicago.Marshall, J. (1956). The Hunters (film). Center for Documentary Anthropology, Somerville,Mass.Marshall, J. (1957). Ecology of the !Kung Bushmen of the Kalahari. Senior honors thesis inanthropology, Harvard University.Marshall, J. (1958). Man as a hunter, parts I and II. Natural History (June-July), 291-309;(August-September), 376-395.Marshall, L. (1957). The kin terminology system of the !Kung Bushmen. Africa 27: 1-25.

Marshall, L. (1960). !Kung Bushman bands. Africa 30: 325-355.Marshall, L. (1961). Sharing, talking and giving: Relief of social tensions among !KungBushmen. Africa 31: 231-249.Marshall, L. (1965). The !Kung Bushmen of the Kalahari Desert. In Gibbs, J. (ed.), Peoplesof Africa. Holt, Rinehart and Winston, New York.Mauss, M. (1906). Essai sur les variations saisonieres des societes eskimos: Etude demorphologie sociale. L'Annee Sociologique 9: 39-132.Meggitt, M. J. (1962). Desert People: A Study of the Walbiri Aborigines of CentralAustralia. Angus and Robertson, Sydney.Piddocke, S. (1965). The potlatch system of the southern Kwakiutl: A new perspective.Southwestern Journal of Anthropology 21: 244-264.Radcliffe-Brown, A. R. (1930). The social organization of Australian tribes, part I. Oceania1: 34-63.Rasmussen, K. (1931). The Netsilik Eskimos. Report of the Fifth Thule Expedition,1921-24, vol. 8. Copenhagen.Schrire, C. (formerly C. White) (in press). The ethno-archaeology of subsistence behaviour in


24/24

rb lee 1972 kung spatial orgn ecol & historical perspective

Documents