c a s t i n g t h e n et w ide - university of...

Oxbow BooksOxford and Oakville

C A S T I N G T H EN E T W I D EPAPERS IN HONOR OF GLYNN ISAAC AND

HIS APPROACH TO HUMAN ORIGINS RESEARCH

Edited by Jeanne Sept and David Pilbeam

JSDP interior Dec052011:Ox template Dec1107 draft 12/7/11 9:22 AM Page v

Published by Oxbow Books on behalf of the American School of Prehistoric Research.

All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical,

photocopying, recording or otherwise, without prior written permission from the publisher.

© Oxbow Books and the individual contributors 2011

Library of Congress Cataloging-in-Publication Data

Casting the net wide : papers in honor of Glynn Isaac and his approach to human origins research /edited by Jeanne Sept and David Pilbeam. -- 1st ed.

p. cm. -- (American school of prehistoric research monograph series)Includes bibliographical references.ISBN 978-1-84217-454-81. Human beings--Origin. 2. Human evolution. 3. Isaac, Glynn Llywelyn, 1937-1985. I. Sept,

Jeanne. II. Pilbeam, David R. GN281.C377 2011599.9--dc23

2011041965

TYPESET AND PRINTED IN THE UNITED STATES OF AMERICA

JSDP interior Dec052011:Ox template Dec1107 draft 12/7/11 9:22 AM Page vi

Contents

Dedication iiiPreface by David Pilbeam ixAcknowledgments xiContributor List xiiiList of Figures and Tables xv

1 Olorgesailie – Retrospective and Current Synthesis: Contributionto the Commemorative Volume in Honor of Glynn IsaacRichard Potts

2 Conversations with Glynn’s Ghost: The Evolution of Paleolandscape Research at East TurkanaAnna K. Behrensmeyer

3 Glynn Llywelyn Isaac: A Recollection and Retrospective AssessmentDiane Gifford-Gonzalez

4 Factors Affecting Variability in Early Stone Age Lithic Assemblages:Personal Observations from Actualistic StudiesNicholas Toth and Kathy Schick

5 Archaeology of Human Origins: The Contribution of West Turkana (Kenya)Hélène Roche

6 The Empire of the Acheulean Strikes BackJohn A. J. Gowlett

7 The Diet of Early Humans: A Summary of the Critical Arguments 40 Years after Isaac’s Initial InsightsManuel Domínguez-Rodrigo

8 Glynn Isaac: Memories and TributeJohn D. Speth

9 Honey and Fire in Human EvolutionRichard W. Wrangham

1

21

41

53

75

93

115

133

149

JSDP interior Dec052011:Ox template Dec1107 draft 12/7/11 9:22 AM Page vii

v i i i Casting the Net Wide

10 A Worm’s Eye View of Primate BehaviorJeanne Sept

11 A Geographic Overview of Neanderthal-Modern Human EncountersOfer Bar-Yosef

12 Sorting Out the Muddle in the Middle East: Glynn Isaac’s Methodof ‘Multiple Working Hypotheses’ Applied to Theories of HumanEvolution in the Late Pleistocene LevantJohn J. Shea

13 Scatters between the Sites – Farm Shelters, Hunting and Gatheringand the Farming Cycle in West Africa: Lessons for ArchaeologicalDistributionsMerrick Posnansky

14 In Pursuit of the Past: A Consideration of Emergent Issues inAfrican Archaeology in the Last Fifty YearsFrancis B. Musonda

15 The Tortoise and the Ostrich Egg: Projecting the Home BaseHypothesis into the 21st CenturyBrian A. Stewart, John Parkington, and John W. Fisher, Jr.

16 A Very Particular Kind of ArchaeologistBernard Wood

169

193

213

231

241

255

279

JSDP interior Dec052011:Ox template Dec1107 draft 12/7/11 9:22 AM Page viii

Contributor List

ContributorsOfer Bar-YosefDepartment of AnthropologyPeabody MuseumHarvard UniversityCambridge, MA 02138, USAE-mail: [email protected]

Anna K. BehrensmeyerDepartment of PaleobiologyNational Museum of Natural History, MRC 121Smithsonian Institution, Washington, DC20013 USA

John W. Fisher, Jr.Department of Sociology and AnthropologyMontana State UniversityBozeman, Montana 59717, USAEmail: [email protected]

Diane Gifford-GonzalezDepartment of AnthropologyUniversity of California, Santa Cruz1156 High Street, Santa Cruz, CA 95064, USAE-mail: [email protected]

John A. J. GowlettSchool of Archaeology, Classics and Egyptology (SACE)University of LiverpoolLiverpool L69 3GS, UK

Francis B. MusondaSenior Lecturer, Department of HistorySchool of Humanities and Social SciencesUniversity of ZambiaP.O. Box 32379Lusaka, [email protected]

John ParkingtonDepartment of ArchaeologyUniversity of Cape TownPrivate BagRondebosch 7701, South AfricaEmail: [email protected]

Merrick PosnanskyProfessor Emeritus, Anthropology and HistoryUniversity of California, Los AngelesLos Angeles, CA 90095, USAE-mail: [email protected]

EditorsJeanne SeptDepartment of AnthropologyIndiana University701 East Kirkwood Ave.Bloomington IN 47405-7100, USAE-mail: [email protected]

David PilbeamDepartment of Human Evolutionary BiologyPeabody MuseumHarvard UniversityCambridge, MA 02138, USAE-mail: [email protected]

JSDP interior Dec052011:Ox template Dec1107 draft 12/7/11 9:22 AM Page xiii

Richard PottsHuman Origins ProgramNational Museum of Natural History,Smithsonian InstitutionWashington, DC 20013-7012, USA; and Department of Earth Sciences, NationalMuseums of KenyaP.O. Box 40658, Nairobi 00100, KenyaE-mail: [email protected]

Hélène RocheDirecteur de Recherche au CNRSUMR 7055 Préhistoire et TechnologieCNRS – Université Paris Ouest Nanterre Maison René Ginouvès, 21 allée de l’Université92023 Nanterre Cedex, France

Kathy SchickCo-Director, The Stone Age Institute1392 W. Dittemore RoadGosport, IN 47433-9582 USAwww.stoneageinstitute.organdDepartment of AnthropologyIndiana UniversityBloomington, IN USAE-mail: [email protected]

John J. SheaAnthropology Department and Turkana Basin InstituteStony Brook University, Stony Brook, NY11794–4364, USAE-mail: [email protected]

John D. SpethMuseum of Anthropology4013 Museums BuildingUniversity of Michigan, Ann ArborAnn Arbor, Michigan 48109-1079, USA E-mail: [email protected]

Casting the Net Widex i v

Brian A. StewartMcDonald Institute for ArchaeologicalResearchUniversity of CambridgeDowning StreetCambridge CB2 3ER, UKEmail: [email protected]

Nicholas TothCo-Director, The Stone Age Institute1392 W. Dittemore RoadGosport, IN 47433-9582 USAwww.stoneageinstitute.org;andDepartment of AnthropologyIndiana UniversityBloomington, IN USAE-mail: [email protected]

Bernard WoodUniversity Professor of Human OriginsCenter for the Advanced Study of HominidPaleobiologyDepartment of Anthropology, GeorgeWashington University2110 G St. NW, Washington, DC 20052, USA

Richard W. WranghamDepartment of Human Evolutionary BiologyPeabody MuseumHarvard UniversityCambridge, MA 02138, USAE-mail: [email protected]

JSDP interior Dec052011:Ox template Dec1107 draft 12/7/11 9:22 AM Page xiv

15

THE TORTOISE AND THE OSTRICH EGG: PROJECTING THE

HOME BASE HYPOTHESIS INTO THE 21ST CENTURY

Brian A. Stewart, John Parkington, and John W. Fisher, Jr.

IntroductionIn the late 1970s Glynn Isaac was promoting ahome base with food-sharing model to under-stand the residual sites of the Plio-Pleistocene ofeast Africa. The debate was around the compe-tencies of our early ancestors and how thesecould be accessed through imaginative manipu-lation of the archaeological record (Binford1981, 1985; Blumenschine 1986; Bunn 1981;Isaac 1978a, 1978b, 1984; Potts 1984, 1988;Sept 1982; Shipman 1983, 1984). The contem-porary florescence of ethnographic documentson hunter-gatherer sites and spatial layouts wasan important inspirational platform (Binford1978, 1983; Gould 1978, 1980; O’Connell1977, 1987; Yellen 1976, 1977). Many felt thatit was a brave, not to say dangerous, leap back-ward from the 20th century Kalahari to the2 million-year-old Turkana Basin. But we sharewith Glynn and his students the belief that it isthe implementation, not the source, of an ideathat matters. We support the view that ‘eureka’can happen anywhere but it is the excavatedrecord that has the final say. We applaud hisdetermination to integrate insights from experi-ments, from ethnographies, from excavationsand from innovative thinking into reconstruc-tions of what might have happened in the past.

Our feeling is that an important strut in thiskind of imagining of the distant past comesfrom an application in the recent past wherepreservation is usually very good, where analo-gous behaviors are more reliably postulated and

where observations, including aspects of sitestructure, are more densely packed. Kroll andIsaac (1984) explored this possibility in moredetailed mapping of the spatial patterning of arti-facts, features and ‘site furniture’ (Binford 1983),showing the potential for reading the social fromthe spatial. Dunefield Midden or DFM, a very latepre-colonial Stone Age site in the Western CapeProvince of South Africa provides just such anopportunity. Because the most persuasive andpervasive models are derived from Kalaharihunter-gatherer observations, this southernAfrican forager behavioral residue should beamenable to the kind of inspired reconstructionIsaac favored. Here we look at the recognition ofhome base structure and sharing through the lensof a large, carefully excavated site with the preser-vation of a wide range of material evidence.

We have chosen to focus on tortoise bonesand ostrich eggshell fragments because, in addi-tion to being very common and almost alwaysprecisely mapped at DFM, they reflect interest-ing trajectories of objects brought to the site asfood items, partly discarded and partly recycledas artifacts, curated, likely shared among fami-lies and ultimately abandoned. As we haveshown elsewhere (Fisher and Strickland 1989,1991), even briefly occupied hunter-gatherercamps have a life history that compounds thedifficulties of reading behavior from spatiallymapped residues. We can try to follow these use-trajectories through the brief time-span of theDFM occupation(s). We emphasize that DFM

JSDP interior Dec052011:Ox template Dec1107 draft 12/6/11 10:36 AM Page 255

differs in significant ways from ethnographicKalahari San sites: DFM’s coastal setting andabundance of shellfish, a greater variety of ther-mal feature types (hearths, roasting pits, stone-packed hearths, etc.) than seems to be the caseat Kalahari sites, a profusion of seal bones, andso forth. We look to the Kalahari San forhypotheses but not for direct analogues.

Below, we present the spatial patterns of tor-toise and ostrich eggshell remains at DFM, andinvestigate the diverse use-histories and deposi-tional trajectories that underpinned them. Wethen explore the fluid boundary between subsis-tence remains and technological remains bydrawing comparisons between tortoises, ostricheggshells and ceramic vessels at the site. Wehope to showcase the importance of using rela-tively recent, exemplar home bases such asDFM to inform spatial interpretations of moreancient home bases like those which Glynnworked so energetically to understand.

Dunefield Midden: Site OverviewDFM is a series of very briefly occupied andmarginally overlapping campsites located onthe Atlantic coast at Elands Bay in the WesternCape Province, South Africa. It lies about 2 kmnorth of the mouth of the Verlorenvlei River andsome 600 m inland from the shoreline (Figure15.1). The rich archaeological deposits containjust under 2,500 kg of shellfish remains, some18,287 stone artifacts, in excess of 10,000 iden-tified bones and teeth of a variety of mammalsand other vertebrates, 796 ostrich eggshell frag-ments, 408 ostrich eggshell beads, 1,011 ceram-ic sherds, and other materials (Orton 2002;Parkington et al. 1992, 2009; Parkington andFisher 2006; Stewart 2005a, 2005b, 2008,2010, 2011; Stynder 2008; Tonner 2005). Themajority of the occupation of DFM took placeduring a span of about a century between AD

1300–1400. A set of 27 radiocarbon dates fromDFM on charcoal and shell has a narrow tempo-ral distribution, with an average of the 27 cali-brated dates being AD 1354 (Parkington et al.2009; Tonner 2005).

Our excavations exposed 859 m2 of shallowliving debris with many different ashy featuresand highly variable distributions of both artifactsand foodwaste. The location of all materials isknown to at least its 1 m2, and many items werepiece plotted. Various lines of evidence supportour belief that the spatial distribution of materialsretains a high degree of integrity (Parkington et al.2009). We have previously defined hearths, roast-ing pits, processing fires and other facilities(Figure 15.2a), demarcated domestic zones anddump zones and suggested different roles for dif-ferent hearths (Parkington 2006; Parkington et al.2009; Stewart 2008; Tonner 2005). Much of thesite structure has survived and provides the spatialframework for looking at activities such as foodconsumption, sharing and waste disposal(Jakavula 1995; Nilsson 1989; Reeler 1992;Stewart 2008; Tonner 2002, 2005).

We discern a probable domestic area and a‘main dump’ at DFM, which lie roughly parallelto one another on a northwest-southeast axis(Figure 15.2b). The domestic area is delineatedby a linear, arcuate arrangement of hearths thatextends from the site’s northern extremity south-eastward toward the southeastern extremity. Themain dump is situated to the west of the domes-tic area and consists of a dense concentration ofshell, animal bones and artifacts. A number ofhearths and many other ashy features also occurin the dump area; some of these might have beencreated early in the site occupation and subse-quently were covered with dumped materials.Several small concentrations of shell and othermaterials – ‘satellite dumps’ – are situated nearor east of some domestic hearths in the northern

Casting the Net Wide2 5 6


Figure 15.1 Map of South Africa’s west coast with Elands Bay and location of Dunefield Midden (DFM).

2 5 7Tortoise and Ostrich Egg


Figure 15.2 Maps of DFM showing (a) the positions of the various feature types, and

(b) the position of the eastern domestic zone of hearths in relation to the main dump to the west.



Tortoise and Ostrich Egg

half of the site (Parkington et al. 2009; Stewart2008; Tonner 2005; Figure 15.2b).

TortoisesMore than 7,500 dispersed fragmentary andcomplete bones of angulate tortoise (Chersinaangulata) have been recorded at DFM. Theassemblage includes some 5,140 carapace speci-mens, 1,500 plastron specimens and some 920non-shell skeletal elements (limbs, axial ele-ments and crania). This species is extremely com-mon in the near coastal sandveld surroundingthe site. From indications of charred plates onboth plastron and carapace bones, it is clear thattortoises were picked up, brought back to campand cooked by simply placing the living animalon the coals of a fire (see Silberbauer 1981:216;Yellen and Lee 1976:41). The tortoise effectivelycooks in its own juices. From the size of the bonesall animals were fairly small, offering much lessthan 1 kg of edible material.

Carapaces often were made into bowls bysimply removing the marginal plates andsmoothing the lateral sutures of the costalplates. Just over 600 carapace bowl parts wererecovered from DFM, which refit to form a min-imum of 30 bowls and probably closer to 50.Adult tortoise carapaces were typically chosenfor bowls, but some were made from (occasion-ally very young) juvenile individuals, raising thepossibility that some functioned as spoons orcompact bowls for special tasks such as prepar-ing arrow poison (Schapera 1930:94, 132,144). Adult or juvenile specimens may havealso been used as ladles (Silberbauer 1981:72).The anteriorly furrowed nuchal plate wasalways left intact and, whether these bowls wereused in food consumption, preparation or both,may have acted as a spout for drinking or pour-ing. Unmodified carapaces may sometimes alsohave been used as bowls (Lee 1979:152).

All tortoise bones, including carapace, plas-tron and non-shell skeletal elements, are morenumerous in the high density shell zone that weregard as a secondary dump area (the ‘maindump’; carapace: 50.2%, n = 2,249; plastron:69.1% n = 1,106; non-shell skeletal: 59.7%, n =610) than in the hearth-dominated eastern areawe interpret as domestic space. All also showrelatively homogenous distributions throughoutthe main dump, occurring more frequently neardump features than other feature types. In thedomestic area, however, the relative frequencyand spatial configurations of each anatomicalcategory varies tremendously. We begin withnon-shell skeletal elements then move on toplastron and carapace parts. Skeletal elementsin the domestic zone are predominantly distrib-uted in and around hearths: 49% (n = 202) arewithin 1 m of hearths. Limb bones and othernon-shell bones of angulate tortoise are smalland after consumption would have become eas-ily buried through trampling or other verticaldisplacement in DFM’s unconsolidated sandsubstrate (Gifford-Gonzalez et al. 1985). Theirhearth-centered domestic distribution thus pre-sumably reflects primary discard (Schiffer1987) – a drop or drool zone in Binford’s (1983)sense (Figure 15.3a).

Plastron parts occur in the domestic area,but they are less densely concentrated than non-shell skeletal elements in and near hearths.Rather, plastrons show a more patchy distribu-tion in this area, with tight clusters near ashyfeatures we interpret as components of satellitedumps that serviced the domestic hearths(Figure 15.3b). Plastrons have no further useafter consumption and thus, like skeletal bones,their distribution presumably also strictlyreflects discard. However, plastron plates aregenerally much bulkier than non-shell bonesand their upright hypoplastron bridges could

2 5 9


Figure 15.3 Spatial distributions at DFM of different categories of (a–c) tortoise foodwaste,

(d) carapace bowl parts, and (e) carapace bowl refits in relation to the various feature types.



Tortoise and Ostrich Egg 2 6 1


potentially cause injury if left dorsal side-up inzones with high foot traffic. Their close associ-ation with ashy patches in satellite dumpstherefore likely results from efforts to clearthese plates from immediate domestic spaceswhere tortoises were typically cooked and con-sumed, indicating either a toss zone (cf.Binford 1983) or a localized secondary refuseaggregate (Murray 1980; Wilson 1994).

Carapace parts are far more widespread inboth the domestic and dump areas than eithernon-shell skeletal elements or plastron parts,occurring at some frequency in most squares atthe site. In the domestic area, some very highdensity concentrations of carapace (>50 speci-mens/m2) exist in/near both hearths and ashyfeatures of satellite dumps (Figure 15.3c).Indeed, comparable frequencies of carapaceparts located outside of the main dump are sit-uated 1 m from hearths and ashy patches.Carapace plates are more brittle than non-shellbones and plastron plates as shown by theirhigh rate of fragmentation. This, along with thepossibility that fragmented carapace plateswere often trampled into the sand substrate(like non-shell bones), may help explain theirubiquity across the site. The most localized dis-tribution is that of carapace bowl parts (Figure15.3d). Bowls are skewed heavily towards themain dump (65.4%, n = 398) in proportionssimilar to that of plastron parts. The few bowlsin the domestic area exhibit spatial patternsthat diverge markedly from those of eachfood-waste category. Rather than diffuse dis-tributions with localized patches of higherdensity in/near ashy patches and/or hearths,bowl parts in the domestic zone almost exclu-sively form several dense (>20 specimens/m2)clusters concentrated next to several hearths(Features 9, 27, 32, 41 and 56). Indeed, 70.6%(n = 149) of bowls in the domestic area were

recovered within 1 m of hearths, whereas only0.92% (n = 2) were situated within the samerange of ashy patches. These domestic hearth-centered specimens show some of the tightestspatial clustering of any carapace bowl refits, andinclude the only complete bowl and half of thenearly-complete bowls at DFM (Figure 15.3e).Such low dispersion of carapace bowl parts sup-ports primary discard, perhaps at a late enoughoccupational stage to preclude secondary clean-up and disposal onto the dump, or perhaps thetight clustering reflects caching in anticipationof future visits.

The overall distribution of carapace bowlsis especially striking relative to unmodifiedcarapace plates and non-shell skeletal ele-ments, both of which are far more diffuselyscattered throughout DFM (compare Figures15.3a, c–d). This is particularly true of theunmodified carapace assemblage. Non-shellskeletal elements, although not much moreabundant than bowl carapace plates, also showa wider distribution across the site and occur,at some frequency, in and near most hearths inthe domestic area. By contrast, bowls are con-centrated in isolated, high density clusters thatin the domestic zone take the form of tightlyrefitting ‘puddles’ near a very limited numberof hearths. We suggest that this is a directreflection of differences in the life histories ofvarious tortoise body parts, with small dispos-able non-shell skeletal elements casually dis-carded in and around domestic hearths, bulki-er carapace and plastron plates typically takenfurther afield to satellite or main dumps, anduseful carapaces modified into bowls, curated,cached and more formally discarded after use.

Ostrich EggshellsThere are just under 800 mapped fragments ofostrich eggs at DFM, of which at least 70% have so




far been refitted, allowing us to estimate the exis-tence of four to seven once-complete eggs. Therefitted eggs had all been used as water flasks, rec-ognizable from the perforations and rounded rimpieces. It is almost certain that more flasks were inuse than we can detect (they simply did notbreak), and that flasks were transported andcached across the landscape for future use. Only2 km away we have excavated a small portion of asite remarkably similar to DFM with a cache offive unbroken eggs (Jerardino et al. 2009;Parkington 2006). It is possible that ostricheswere not common in this part of the sandveld andthat eggs were systematically brought in as flasksand regularly buried as a source of fresh water.These caches are nearly impossible to detect inthe sandy landscape and are probably far morecommon than we realize.

Eggshell fragments are, by comparison withother artifacts and foodwaste, rather restrictedin their distribution. They are far more commonin the northern half of the site than in the south(Figure 15.4a); in the latter they form smallpatches that often refit with larger clusters in thenorth (the exception is one much larger, diffuseconcentration near the southern hearth Feature63). Most ostrich eggshell fragments were recov-ered from the main dump (71.3%, n = 561). Thisproportion is higher than any of the tortoise cat-egories discussed above, but not far off fromplastron (69.1%, n = 1,106) and bowl parts(65.4%, n = 398). In the main dump, ostricheggshell fragments occur in three major concen-trations in the north part of the site and a fourthin the south (near Feature 63). They associatewith highest frequency near dump features, fol-lowed by processing features and then hearths.The relatively high frequency near processingfeatures is something eggshell fragments sharewith tortoise carapace bowls. The behaviorsresponsible for these associations are obscure

since it is difficult to envisage ostrich eggshellsbeing used in food processing, unless, of course,it was the eggs themselves being prepared.Among Kalahari groups, however, ostricheggshells are never broken during consumption,but rather the egg contents are accessed throughthe same perforation that later serves as the flaskspout (Lee 1979; Silberbauer 1981).

In the domestic area, the distribution ofostrich eggshell fragments recalls tortoise cara-pace bowls in that relatively large concentrationscluster around a few specific hearths (Features10, 11 and 22), with other hearths hosting one ortwo isolated fragments (Figure 15.4a). Thirty onepercent (n = 70) of eggshell fragments outside themain dump are within 1 m of hearths, whereas17.3% (n = 39) occur within 1 m of ashy patches.But unlike the bowls, the specific hearths aroundwhich eggshell fragments concentrate are notoverlain by subsequent dump deposits and so canbe more confidently associated with their adja-cent eggshell scatters. We thus interpret thedomestic distribution of eggshell fragments asindicating that flasks were in use at these hearthsand subsequently discarded when broken.Fragments recovered near hearths presumablyreflect incomplete hearthside maintenance orfragment reuse. Fragments deemed waste werediscarded either in nearby satellite dumps or inthe main dump.

A detailed picture of ostrich eggshell flasksuse, reuse and disposal can be reconstructed fromthe flask refit patterns, which are some of themost interesting of any class of archaeologicalmaterial at DFM (Jakavula 1995; Figure 15.4b).Numerous connections exist within and betweendifferent eggshell clusters, features and areas ofthe site, showing that the flasks underwentremarkably dynamic depositional trajectoriesoften involving long-distance movements. Werecognize three broad types of connection:

2 6 3


Figure 15.4 Spatial distributions at DFM of (a) ostrich eggshell flask fragments, (b) ostrich

eggshell flask refits, and (c–d) ostrich eggshell beads in relation to the various feature types.



1) refits within the domestic area; 2) refitsbetween the domestic area and the main dump;and, 3) refits within the main dump.

Within the domestic zone a web of refitslinks five individual domestic hearths (Features11, 12, 22, 33, and 35; Figure 15.4b). We sug-gest that this reflects the daily movements ofpeople between individual domestic spacesorganized around fireplaces. As with brokenpots (see below), large fragments of brokenostrich eggshell flasks still would have been use-ful as plates, bowls, scoops, spoons, bead blanksand so forth – particularly if, as we suspect,ostrich eggs were difficult to come by in thelocal sandveld landscape. These hearth-to-hearth connections probably thus reflect theextended use-lives of these valuable objects aspeople transported and broke flasks, thenshared and discarded the fragments. That theweb knits together three neighboring hearths(Features 11, 12, and 22) raises the possibilitythat these features belonged to people who wereclose, whether kin relations or friends (cf. Yellen1977). In addition to hearth-to-hearth connec-tions, refits within the domestic zone also implydiscard processes. Two of the domestic hearthslinked by the web discussed above (Features 22and 12) connect to the satellite dump directlynorth of Feature 12 (associated with two ashpatches, Features 85 and 87), which suggeststhat different domestic units (families?) wereusing the same satellite dump, further strength-ening their spatial (and social?) connectedness.

A number of the eggshell clusters involved inthese hearth-to-hearth and hearth-to-satellitedump relationships link into the larger eggshellconcentrations in the main dump (Figure 15.4b).These domestic-to-dump refits provide fairlyunambiguous evidence for secondary discardprocesses; exhausted flask portions were collect-ed during the cleaning of hearthside domestic

spaces or satellite dumps and re-deposited in thesite’s primary waste accumulation, the maindump. But they also give a sense of how the maindump formed, with (what we presume to havebeen) individual domestic units consistentlydepositing their waste in specific (prescribed?)parts of the dump, creating initially separate mid-dens that coalesced with time (Parkington et al.2009:116). The refit patterns show that theseindividual dumps were located in areas of themain dump nearest to the living spaces of thegiven domestic units that produced them.

Finally, there are refits between ostricheggshell fragments within the main dump. Themain dump includes four major concentrationsof eggshell fragments, each exhibiting consider-able internal connectivity in refits (Figure15.4b). However, occasional refits also link theconcentrations to one another, which indicatesthey are contemporary, and again supports thenotions that the main dump served interlinkeddomestic units and consists of several separatedumps that eventually fused together. The dis-tances over which eggshell fragments refit in themain dump are extremely variable, rangingfrom a few centimeters to over 18 m (Figure15.4b). These attenuated conjoins show thatalthough discarded eggshells at DFM are inti-mately bound up with domestic processes in thenorth, occasional activities brought fragmentsfurther south, echoing the north-south connec-tivity exhibited by the domestic hearth-to-hearth refit patterns.

The above distributional and refit patterninggives an idea of ostrich eggshell flask use, reuse,sharing and discard at DFM. Also present at thesite are ostrich eggshell beads which are, ofcourse, another form of flask fragment reuse.Beads represent the far ‘distal end’ of an ostrichegg’s use-life. Interestingly, broken but finishedbeads are more common in the dump, and




whole finished ones are more common nearhearths (Figure 15.4d). But despite the evidencepresented above for hearth-to-hearth reuse ofbroken flasks, little attempt appears to havebeen made at DFM to turn ostrich eggshell frag-ments into beads. Out of a total of 408 ostricheggshell beads at DFM, only three are unfin-ished and none were abandoned or broken inthe course of manufacture (Kandel and Conard2005; Orton 2008).

Spatially, beads and flask fragments consis-tently occur in different areas of both thedomestic zone and especially the main dump(Figure 15.4a, c). In the domestic zone, thebead distribution is heavily skewed toward thesouth where several large clusters center direct-ly on hearths. This pattern is the reverse of thatexhibited by eggshell flask fragments; whereasthe southern domestic area is virtually devoid offlask fragments, the north contains several size-able concentrations. In the main dump thesetwo artifact types coincide more frequently, butagain their distributions diverge. From Figure15.4a and 15.4c it is clear that squares withbeads often occupy the spaces ‘between’ thosewith flasks fragments. As with the domesticzone, most beads in the main dump are situatedin the south. This, again, contrasts sharply withthe distributions of flask fragments, the vastbulk of which comprise one of three majormain dump concentrations in the north. Thisspatial mismatch of ostrich eggshell beads andflask fragments suggests that these artifactsexperienced independent depositional trajecto-ries. This is consistent with relatively minimalevidence for on-site bead production, indicatingthat although the use-lives of the ostricheggshell flasks may have extended beyondbreakage, they were not attenuated enough, atleast at DFM, for fragments to be recycled intowholly new artifact forms.

Comparing Containers: Carapace Bowls,Eggshell Flasks and Ceramic VesselsCeramic vessels constitute a third container typeemployed by the DFM inhabitants in addition totortoise carapace bowls and ostrich eggshellflasks (see Stewart 2005a, 2005b). Unlike thelatter two, however, pots never underwent acomparable phase of being a food resource priorto becoming an artifact, and thus serve as a use-ful ‘yardstick’ against which to compare the spa-tial distributions of bowls and flasks. Here, weexpand our view to consider the life histories ofthese three container types, and contemplatehow much of that life history transpired at DFMand was captured in this site’s archaeologicalrecord. Because they differed sharply from oneanother in raw material availability, manufactur-ing effort, size, robusticity and, very probably,perceived value, our expectation is that the lifehistories of flasks, bowls and pots were diverse.Among our considerations is the notion that thenumber of specimens deposited at DFM andtheir spatial distribution might have been condi-tioned by the ‘turnover rate’ of each artifact type.

Tortoise carapace bowls began their artifactlife history in the landscape outside of DFMwhen they were collected, most likely as live tor-toises, and subsequently brought to DFM. AtDFM the tortoises presumably were cooked andconsumed, and afterward some carapaces weremade into bowls. We suggest that carapacebowls typically spent most or all of their use-lifeat DFM. They could be replaced easily, so long asthe availability of living tortoises remained ade-quate. Thus the turnover rate of carapace bowlsmight have been relatively rapid. If the DFMoccupants used carapace bowls for serving foodand eating, as in the Kalahari, they were likely toenter the archaeological record at DFM ratherthan at an offsite location. A high turnover ratecoupled with discard at DFM rather than offsite

2 6 7


Figure 15.5 Spatial distributions at DFM of (a–b) tortoise carapace bowl parts and refits, (c–d) ostrich eggshell flask

fragments and refits, and (e–f) ceramic sherds and refits in relation to the domestic zone of hearths and the main dump.



would result in a relatively high frequency ofspecimens; this is indeed the case, with 30bowls a minimum estimate for DFM andupwards of 50 more likely.

Carapace bowl refits occur almost exclusive-ly within the dump (Figure 15.5b). The few thatoccur in the domestic zone never connect two ormore domestic hearths but instead form tight‘puddles’ of near-complete bowls near severalindividual domestic hearths, and only one bowlconnects the domestic area with the maindump. Moreover, the refit distances of carapacebowls within the dump span much shorter dis-tances than is the case for ostrich eggshell andceramic fragments. The combined evidence sug-gests that carapace bowls were rendered uselessquickly after breakage, perhaps once one or twoplates had fallen out, and disposed of in thedump while still relatively intact rather thanafter fragments had become dispersed. This isconsistent with our notion that the turnover rateof these artifacts was relatively rapid.

We suggest that the life history of ostricheggshell as an artifact differed significantly fromtortoise carapace bowls in that ostrich eggs typi-cally had lower turnover rates and generally‘passed through’ DFM, whereas bowls were some-thing more akin to ‘permanent residents’ at thesite. Ostrich eggs also began their artifact life his-tory away from DFM when they were obtainedsomewhere in the landscape. If ostriches andtheir eggs were rare in the sandveld, efforts wouldhave been made to prolong their use-lives (Lee1979:274). Also, if use as water flasks by the DFMoccupants was the principal function of ostricheggs, we suggest that flasks spent much of theirlife away from DFM. Flasks presumably wouldhave accompanied women and men on gatheringand hunting expeditions, respectively (Lee 1979;Silberbauer 1981). Further, the absence at DFMof whole flasks or even large fragments suggests

those that remained intact when the inhabitantsmoved on were taken from DFM. The upshot of allthis is that most flasks whose artifact life ended inbreakage probably came to this end away fromDFM. DFM was just a “stop” in what was a geo-graphically broad network of use, and few flasksentered the archaeological record there. This sce-nario is consistent with the small number of flasksat the site – five to seven specimens.

Flask refits, unlike those of tortoise carapacebowls, connect between different domestichearths with some regularity (Figure 15.5d).Perhaps broken flask fragments had useful func-tions as implements of some kind (e.g., plates,spoons, scoops), and once a flask broke thepieces were shared among multiple hearths(households?). We submit that the most ‘intu-itive’ reason for sharing broken ostrich eggshellfragments is that they provide raw material forostrich eggshell bead manufacture. But there isconsiderable evidence to suggest that the siteoccupants engaged only minimally in bead pro-duction at DFM. For now we lack a compellingexplanation for this pattern, but the limitednumber of water flasks at DFM combined withtheir long refit distances is consistent with theidea that ostrich eggshells represented rare andthus relatively precious artifacts to be handledcarefully, curated, perhaps reused as implements(if not recycled as beads) or, if still whole, trans-ported away from DFM.

Ceramic vessels, of which there are at least20 at DFM, also started their life history offsitewhere they were manufactured or acquiredthrough trade with neighboring ceramic-produc-ers (Stewart 2005b, 2008). Since the DFM potswere clearly employed in cooking (Stewart2005a), their life history, like that of carapacebowls, was probably more strongly centered thanostrich eggshell flasks at DFM than at offsite loca-tions. But similar to flasks (and unlike carapace




bowls), complete pots or serviceable pot por-tions would have been transported offsite whenthe occupants moved. Whether the pots weremanufactured by the DFM occupants or tradedin, the ceramic assemblage represents a substan-tial investment in time and/or resources. Wesuggest that efforts to protect, repair and reusepots (Stewart 2008), along with their relativelyhigh robusticity, would have resulted in use-livesthat were certainly longer than carapace bowlsand perhaps also ostrich egg flasks. But we alsoexpect that the greater amount of time potsspent at DFM relative to flasks would haveoffered more opportunities for breakage on-site,resulting in more vessels than flasks but perhapsfewer than carapace bowls (which, though also‘site-based’, were clearly more expendable). Thisis indeed what we see; DFM contains at most 25ceramic vessels, more than the maximum num-ber of ostrich egg flasks (n ! 7) and less thancarapace bowls (n ! 50).

The distributions of ceramic sherds (Figure15.5e) echo those of ostrich egg flasks and tor-toise carapace bowls in that each of these classesof artifact shows spatially tight concentrations ofsmall numbers of specimens adjacent to andclearly associated with domestic hearths. Theseclusters might represent specimens that were bro-ken shortly before the occupants departed DFMand thus were not cleaned up onto the dump, orthey could be cached vessels (Nilssen 1989;Stewart 2008). Relative to ostrich eggshell flasks,the ceramic refits infrequently connect separatedomestic hearths, perhaps implying that ceramicvessels rarely moved between domestic house-holds. But the ceramics exhibit a much moreexpansive network of refits than either ostrich eggflasks or carapace bowls, with frequent connec-tions between the domestic area and the maindump and within the dump (Figure 15.5f). Wesuggest that the relatively long use-lives and the

site-based life histories of the DFM pots arereflected in these refit patterns. The use-life tra-jectory of ceramic vessels harbors potentiallysubstantial complexity, and broken specimenscould still be functional, albeit for uses differentfrom their original function (Stewart 2005b,2008). A range of stages in the use-life of vesselsis represented by specimens of varying com-pleteness, and different pieces of the same vesselare sometimes distributed among various loca-tions at DFM, probably reflecting a “sequence”through time of a vessel being reduced by break-age and pieces re-used then eventually discard-ed (Stewart 2008).

Discussion: From Foodwaste to ArtifactMaking sense of the spatial distributions of tor-toise remains and ostrich egg fragments at DFMinvolves a consideration of seemingly complex lifehistory trajectories, inasmuch as these two itemsoffered the DFM inhabitants both a meal and araw material from which to manufacture artifactsranging from the practical to the ornamental. Wefind ourselves striving to integrate a diverse set offactors as we attempt to disentangle these differ-ing but interconnected pathways. The factorsinvolved range from cooking, consumption, anddiscard patterns conditioned by the size andstructure of the food “package” to the trajectoryand location of artifact manufacture, use, and dis-card to the “nuisance value” (O’Connell 1995) ofobjects and how that influenced the manner andlocation of disposal. Here we discuss severalimplications of these complex foodwaste-to-arti-fact trajectories for interpreting archaeologicalspatial patterning both at DFM and at older, lesswell preserved archaeological home bases.

A first observation is that the difference inavailability of a food resource can directly affectthe value of an artifact made from its wastebyproducts and, by extension, the artifact’s spatial

2 7 1


distribution at home bases in the archaeologicalrecord. Tortoises abound in DFM’s sandveldlandscape. Ample access to these animalsreduced the value of carapace bowls. Even if asingle carapace plate of a bowl fell out or broke,another bowl could be rapidly produced from analmost inexhaustible supply of live tortoises inthe landscape or from discarded carapaces atDFM. Thus, carapace bowls at DFM probablyhad high turnover rates and negligible reuse offragments upon breakage. The result is thatalthough carapace bowls are the most abundantcontainer type at DFM, they exhibit the shortestrefit distances. This foodwaste-to-artifact path-way contrasts sharply to that of ostrich eggs,which are far less frequent in the sandveld land-scape than tortoises. The limited accessibility ofostrich eggs and their importance as waterflasks lowered their turnover rates and raisedfrequencies of flask fragment reuse.Consequently, there are far fewer ostrich eggflasks relative to carapace bowls at DFM, butindividual flasks are widely dispersed with websof long-distance refits linking domestic hearthswith one another and with the main dump.

The analysis of spatial patterns of items thatcould be both foodwaste and artifact also hasmethodological implications for transcendingthe distinction between subsistence and tech-nology at archaeological home bases. Strikingdifferences in the spatial distribution of tortoisecarapace bowls compared to unmodified cara-pace at DFM offers perhaps the strongest signalseparating foodwaste and artifact. Carapacebowls occur largely in the dump whereasunmodified carapace occurs all throughout thedump and the domestic area, where they oftenassociate with hearths. This dissimilarity sug-gests a significant difference in disposalbetween artifact and foodwaste – when it cametime to dispose of carapace bowls they usually

went to the dump, while disposal of unmodifiedcarapace was more widespread across the site.The progressive disintegration and dispersalthrough time of a discarded, unmodified cara-pace or plastron into individual plates in thedomestic area could result in rather widespreaddistribution of fragments. But could the differ-ences in carapace distribution be detected atarchaeological sites of people, such as theJu/’hoansi, who did not modify carapace bowlsbefore using them? If not, these important arti-facts and their use-contexts would be misinter-preted as strictly subsistence-related.Archaeologists must be able to distinguish thespatial patterning of artifact and foodwaste cate-gories, even when they are manifested by speci-mens made of the same raw material. We havepresented steps in this direction in our analysesof tortoise and ostrich egg at DFM.

Finally, the foodwaste/artifact dichotomyholds implications for discerning social relation-ships and organization at archaeological homebases. Most efforts to reconstruct social connec-tivity at prehistoric campsites have focused onsubsistence remains, particularly those of largeanimals that would have been shared betweendomestic units. The success of this approachdepends directly on these bones being in pri-mary context, such as near hearths aroundwhich consumption took place, which is mostlikely to happen at sites where occupation spanswere short. But ethnoarchaeological researchdemonstrates that prehistoric home bases occu-pied for a longer period of time are more likelyto have survived for archaeologists to detectthem and, unfortunately, longer occupationaldurations mean more intensive secondary dis-card processes. DFM is a case in point. Here,larger animals such as eland and steenbok aremore likely to have been shared between socialunits such as nuclear families, but also are more




likely to have ended up on the dump where thesocial connections are harder to trace (Stewart2008). Smaller animals such as tortoise aremore likely to have remained in primary con-text, but are less likely to have been sharedbetween social units. By contrast, although someartifact types at DFM, notably carapace bowls,generally ended up on the dump, ceramics andostrich egg flasks occur in the domestic area,and refits connect separate hearths. It might be,therefore, that artifacts – and especially thosethat were curated and shared – hold greaterpotential than subsistence remains for recogniz-ing social connections at many archaeologicalhome bases since efforts to remove them fromdomestic contexts were less rigorous.

Final ReflectionsThe objective of Glynn Isaac’s pursuit of a homebase model in the 1970s was to demonstrate thatearly Pleistocene sites could plausibly be inter-preted as home bases – places to which ancientmen and women returned with food items sepa-rately gathered. The ‘home base’ concept impliedthe existence of a number of behavioral associa-tions, including a division of labor, the capacity toagree on reuse of specific places as central loca-tions from which to forage, and food sharing. InIsaac’s view, early Pleistocene foragers could beshown already to exhibit behavioral patterns welldocumented in the ethnographic record. Otherarchaeologists, of course, preferred to see a muchlater appearance of such ‘modernity’ (and Isaachimself modified his views). Despite this, theobjective was rather simple. Positing the existenceof a home base needed, according to Isaac, onlythe demonstration that faunal assemblages docu-mented the residues of multiple episodes of ani-mal food acquisition.

Since then ethnoarchaeologists and archae-ologists have adopted much more ambitious

agendas for the interpretation of scatters andpatches of apparently associated faunal and arti-factual remains. At DFM, for example, we recog-nize that a primary obligation is to persuasivelyargue for brevity of occupation and reliability ofassociation as the underpinnings for a numberof further inferences and conclusions. A large,reliably associated assemblage that includesfood debris and artifactual wastes deriving frommany sources and materials provides invaluableevidence as to the identity of occupants, the sea-sonality of visits, the exploitation of resources andthe domestic organization of groups. Included inthe behavioral objectives are the recognition anddocumentation of food and artifact sharing,waste disposal, artifact manufacture and gen-dered uses of features and spaces. Comparing themapped distributions of similar and different setsof items, and tracing their variable trajectoriesusing refits, lets us hypothesize about the behav-ioral dynamics that suffused and structuredhome bases in prehistory. Because even a huntergatherer camp has a life history, it becomes pos-sible to try to trace the passage of objects into,through and out of briefly occupied sites.

From our discussion above of tortoise andostrich eggshell remains, it is obvious that there isa deep level of connectivity across the 859 m2 wehave excavated. Links between features that revealcontemporaneity of occupation across the wholeexcavated space boost our confidence in the brevi-ty of occupation, and increasingly support thetight radiocarbon chronology. Dissecting this con-nectivity, we have argued for some specific uses offeatures within a broad distinction betweendomestic and waste disposal zones, with somefeatures associated with particular types of arti-facts and/or non-artifactual animal remains andothers not, some quite focused, others less so. Wehave examined in some detail two such remains –tortoise and ostrich eggshell – that are particularly

2 7 3


References

Binford L. R. 1978 Nunamiut Ethnoarchaeology. Academic Press, New York.

1981 Bones: Ancient Men and Modern Myths. Academic Press, New York.

1983 In Pursuit of the Past: Decoding the Archaeological Record. Thames and Hudson, New York.

revealing since both often transcended the porousboundary between subsistence and technology.We have argued that: 1) the availability of a foodresource directly influences the abundance andspatial patterning of the artifactual byproduct;2) interpretations of spatial patterning depend onthe ability to distinguish between foodwaste andartifacts; and, 3) artifacts might yield better infor-mation than foodwaste about social connectivityat many home bases. We feel this last point to beparticularly salient.

The success of spatial research at archaeologi-cal home bases in the 21st century depends onour capacity to bridge the yawning temporal andinferential gap between 20th century hunter-gath-erers and archaeological traces of far more ancientforagers. The analytical power of relatively recentsites like DFM lies in the richness, diversity andpreservation of remains, which greatly enhancesthe strictly ‘stones and bones’-based archaeologythat typifies most Paleolithic enquiries. But even atsites like DFM interpretive challenges loom.Efforts must be made to recognize the early,patchy accumulation of dumps that subsequentlyexpanded and coalesced, swamping some hearthswith their associated debris. More detailed workon tortoise limb bone distributions will likely firmup the recognition of consumption zones. Thatminimal bead making took place may have to dowith the vagaries of behaviors on briefly occupiedsites, but accurate interpretation must wait until

other local and regional comparisons becomeavailable. These future directions should furtherstrengthen our inferences at DFM, and with themour confidence that recent, well preserved andextensively excavated home bases like DFM canserve as stepping stones back into earlier times ofthe Paleolithic.

AcknowledgementsWe thank Dolores Jacobs, whose indefatigablework in processing and refitting archaeologicalmaterials from DFM underlies much of what wehave accomplished here. A good deal of what wehave presented in this paper builds on earlierrefitting and/or recording work by ZukisaniJakavula, Peter Nilssen and Tobias Tonner, all ofwhom we gratefully acknowledge. We are alsograteful to Stuart Challender for providing GISexpertise. B. A. S thanks Oxford University Pressand St. Hugh’s College, Oxford, for doctoralfunding. J. P. thanks the University of Cape Townfor sustained logistical and financial support. J.W. F. thanks Montana State University for aScholarship and Creativity grant that supportedhis research on DFM materials. We thankJeanne Sept and David Pilbeam very deeply forinviting us to contribute to this memorial forGlynn. Above all we acknowledge Glynn Isaac,who with warm and cheerful countenance ori-ented Stone Age archaeology in directions thatcontinue to remain so relevant today.




1985 Human ancestors: Changing views of their behavior. Journal of Anthropological Archaeology 4:292–327.

Blumenschine, R. J. 1986 Carcass consumption sequences and the archaeological distinction of scavenging and hunting. Journal of

Human Evolution 15:639–659.

Bunn, H. T. 1981 Archaeological evidence for meat-eating by Plio-Pleistocene hominids from Koobi Fora and Olduvai Gorge.

Nature 291:574–577.

Fisher, Jr., J. W., and H. C. Strickland1989 Ethnoarchaeology among the Efe Pygmies, Zaire: Spatial organization of campsites. American Journal of

Physical Anthropology 78:473–484.

1991 Dwellings and fireplaces: Keys to Efe Pygmy campsite structure. In Ethnoarchaeological Approaches to Mobile

Campsites, edited by C. S. Gamble and W. A. Boismier, pp. 215–236. International Monographs in Prehistory, Ann

Arbor.

Gifford-Gonzalez, D., D. B. Damrosch, J. Pryor, and R. L. Thunen1985 The third dimension in site structure: An experiment in trampling and vertical dispersion. American

Antiquity 50:803–818.

Gould, R. A. (Editor) 1978 Explorations in Ethnoarchaeology. University of New Mexico Press, Albuquerque.

Gould, R. A. 1980 Living Archaeology. Cambridge University Press, Cambridge.

Isaac, G. L. 1978a The food sharing behaviour of protohuman hominids. Scientific American 238:90–106.

1978b Food sharing and human evolution: Archaeological evidence from the Plio-Pleistocene of East Africa.

Journal of Anthropological Research 34:311–325.

1984 The archaeology of human origins: Studies of the Lower Pleistocene in East Africa 1971–1981. Advances in

World Archaeology 3:1–97.

Jakavula, Z. 1995 More Than Lines on a Map?: Refitting Dunefield Midden. Unpublished Honours thesis, Department of

Archaeology, University of Cape Town.

JSDP interior Dec052011:Ox template Dec1107 draft 12/6/11 10:36 AM 5

Jerardino, A., L. K. Horwitz, A. Mazel, and R. Navarro 2009 Just before Van Riebeck: Glimpses into terminal LSA lifestyle at Connies Limpet Bar, West Coast of South

Africa. South African Archaeological Bulletin 64:75–86.

Kandel A. W., and N. J. Conard 2005 Production sequence of ostrich eggshell beads and settlement dynamics in the Geelbek Dunes of the

Western Cape, South Africa. Journal of Archaeological Science 32:1711–1721.

Kroll, E. M., and G. L. Isaac1984 Configurations of artifacts and bones at Early Pleistocene sites in East Africa. In Intrasite Spatial Analysis in

Archaeology, edited by H. Hietala, pp. 4–31. Cambridge University Press, Cambridge.

Lee, R. B. 1979 The !Kung San: Men, Women and Work in a Foraging Society. University of Cambridge Press, Cambridge.

Murray, P. 1980 Discard location: The ethnographic data. American Antiquity 45:490–502.

Nilssen, P. J. 1989 Refitting Pottery and Eland Body Parts as a Way of Reconstructing Hunter-Gatherer Behavior: An Example

from the Later Stone Age at Verlorenvlei. Unpublished Honours thesis, Department of Archaeology, University of

Cape Town.

O’Connell, J. F. 1977 Ethnoarchaeology of the Alyawarra: A report. AIAS Newsletter 7:47–49.

1987 Alyawara site structure and its archaeological implications. American Antiquity 52:74–108.

1995 Ethnoarchaeology needs a general theory of behavior. Journal of Archaeological Research 3:205–255.

Orton J. 2002 Patterns in stone: The lithic assemblage from Dunefield Midden, Western Cape, South Africa. South African

Archaeological Bulletin 57:31–37.

2008 Later Stone Age ostrich eggshell bead manufacture in the Northern Cape, South Africa. Journal of

Archaeological Science 35:1765–1775.

Parkington, J. E. 2006 Shorelines, Strandlopers and Shell Middens. Krakadouw Trust, Cape Town.




Parkington, J. E., and J. W. Fisher, Jr.2006 Small mammal bones on Later Stone Age sites from the Cape (South Africa): Consumption and ritual

events. In Integrating the Diversity of Twenty-First-Century Anthropology: The Life and Intellectual Legacies of Susan Kent,

edited by W. Ashmore, M.-A. Dobres, S. M. Nelson, and A. Rosen, pp. 71–79. Archaeological Papers of the

American Anthropological Association, Vol. 16.

Parkington, J. E., J. W. Fisher, Jr., and T. W. W. Tonner 2009 ‘The fires are constant, the shelters are whims’: A feature map of Later Stone Age campsites at the Dunefield

Midden site, Western Cape Province, South Africa. South African Archaeological Bulletin 64:104–121.

Parkington, J. E., P. Nilssen, C. Reeler, and C. Henshilwood1992 Making sense of space at Dunefield Midden campsite, western Cape, South Africa. Southern African Field

Archaeology 1:63–70.

Potts, R. 1984 Home bases and early hominids. American Scientist 72:338–347.

1988 Early Hominid Activities at Olduvai. Aldine de Gruyter, New York.

Reeler, C. 1992 Spatial Patterns and Behaviour at Dunefield Midden. Unpublished M.A. thesis, Department of

Archaeology, University of Cape Town.

Schapera, I. 1930 The Khoisan Peoples of South Africa: Bushmen and Hottentots. Routledge and Kegan Paul, London.

Schiffer, M. B. 1987 Formation Processes of the Archaeological Record. University of New Mexico Press, Albuquerque.

Sept, J. 1982 Was there no place like home? Current Anthropology 33:187–207.

Shipman, P. 1983 Early hominid lifestyle: Hunting and gathering or foraging and scavenging? In Animals and Archaeology:

Volume I, Hunters and Their Prey, edited by J. Clutton-Brock and C. Grigson, pp. 31–49. British Archaeological

Reports, Oxford.

1984 Ancestors: Scavenger hunt. Natural History 93:20–27.

Silberbauer, G. B. 1981 Hunter and Habitat in the Central Kalahari Desert. Cambridge University Press, Cambridge.


Stewart, B. A. 2005a Charring patterns on reconstructed ceramics from Dunefield Midden: Implications for Khoekhoe vessel

form and function. Before Farming [online version] 2005/1 article 1.

2005b The Dunefield Midden ceramics: Technical analysis and placement in the Western Cape sequence. South

African Archaeological Bulletin 60:103–111.

2008 Refitting Repasts: A Spatial Exploration of Food Processing, Sharing and Disposal at the Dunefield Midden

campsite, South Africa. Unpublished D.Phil. Dissertation, Institute of Archaeology, University of Oxford.

2010 Modifications on the bovid bone assemblage from Dunefield Midden, South Africa: Stage one of a multi-

variate taphonomic analysis. Azania: Archaeological Research in Africa 45:238–275.

2011 ‘Residues of parts unchewable’: Stages two and three of a multivariate taphonomic analysis of the Dunefield

Midden bovid bones. Azania: Archaeological Research in Africa 46:141–168.

Stynder, D. D. 2008 The impact of medium-sized canids on a seal bone assemblage from Dunefield Midden, West coast, South

Africa. South African Archaeological Bulletin 63:159–163.

Tonner, T. W. W. 2002 A spatial database for the Later Stone Age site ‘Dunefield Midden’ (Western Cape, South Africa).

Unpublished M.A. thesis, Department of Archaeology, University of Cape Town.

2005 Later Stone Age shellfishing behaviour at Dunefield Midden (Western Cape, South Africa). Journal of

Archaeological Science 32:1390–1407.

Wilson, D. C. 1994 Identification and assessment of secondary refuse aggregates. Journal of Archaeological Method and Theory

1:41–68.

Yellen, J. E. 1976 Settlement patterns of the !Kung: An archaeological perspective. In Kalahari Hunter-Gatherers: Studies of the !Kung

San and their Neighbors, edited by R. B. Lee and I. DeVore, pp. 47–72. Harvard University Press, Cambridge (MA).

1977 Archaeological Approaches to the Present: Models for Reconstructing the Past. Academic Press, New York.

Yellen, J. E., and R. B. Lee 1976 The Dobe-/Du/da environment: Background to a hunting and gathering way of life. In Kalahari Hunter-

Gatherers: Studies of the !Kung San and their Neighbors, edited by R. B. Lee and I. DeVore, pp. 27–46. Harvard

University Press, Cambridge (MA).



c a s t i n g t h e n et w ide - university of...

Documents