8/22/2019 Schiffer 72

1/11

Society for American Archaeology

Archaeological Context and Systemic ContextAuthor(s): Michael B. SchifferReviewed work(s):Source: American Antiquity, Vol. 37, No. 2 (Apr., 1972), pp. 156-165Published by: Society for American ArchaeologyStable URL: http://www.jstor.org/stable/278203 .

Accessed: 20/05/2012 15:04

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact support@jstor.org.

Society for American Archaeology is collaborating with JSTOR to digitize, preserve and extend access to

American Antiquity.

http://www.jstor.org

http://www.jstor.org/action/showPublisher?publisherCode=samhttp://www.jstor.org/stable/278203?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/278203?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=sam

8/22/2019 Schiffer 72

2/11

ARCHAEOLOGICAL CONTEXT AND SYSTEMIC CONTEXTMICHAELB. SCHIFFER

ABSTRACTThe cultural aspect of the processes responsible for forming the archaeological record is argued to be anunderdeveloped branch of archaeological theory. A flow model is presented by which to view the "life history"or processes of systemic conte of any material element. This model accounts for the production of asubstantial portion of the archaeological record. The basic processes of this model are: procurement,manufacture, use, maintenance, and discard. Refuse labels the state of an element in archaeological context. Thespatial implications of the model suggest a largely untapped source of behavioral information. Differential refusedisposal patterns are examined as they affect artifact location and association. The meaning of element relativefrequencies in refuse is discussed.Department of AnthropologyUniversity of ArizonaMay, 1971

Perhaps the most important assumption made by many archaeologists s that the spatialpatterningof archaeologicalemainsreflectsthe spatialpatterningof pastactivities Binford1962,1964; Brose 1970; Clarke1968; Hill 1970a, 1970b; Longacre1970; McPherron1967; Struever1968;Wilmsen1970;andmany others).The loss, breakage,and abandonmentof implementsand facilities at different ocations,wheregroupsofvariabletructureerformedifferentasks, eavesa "fossil"ecord f the actual perationf an extinctsociety [Binford1964:4251.

This statement suggests that the proveniencesof artifactsin a site correspondto their actuallocations of use in activitvities.learlythis is not always the case. But to what extent it may be soandunder what conditions,would appearto be a problemworthyof some attention. In the mostgeneralterms, what I am askingis, How is the archaeological ecord formed by behavior in aculturalsystem?I would emphasizethat I am not asking the equally generaland importantquestionof whythere is variabilityin past cultural systems. I amn quiringfirst,Whyis there an archaeologicalrecord?Second, How does a cultural system produce archaeologicalremains?And third,Whatkinds of inter-culturaland intra-cultural ariablesdeterminethe structure(as distinct from theform andcontent) of the archaeologicalecord?The branchof archaeologicalheory whichtreatsthese and relatedquestionsmaybe defined asthe conceptual system that explainshow the archaeologicalecord s formed. As such,it hasbothculturaland non-culturallturalcomponents.Thelatter areahas receivedmajoremphasis o date;regularpatternsof post-depositionalchanges n both artifactinventoriesand site morphologieshavebeenelucidated(Hole and Heizer 1969). Explanationsof variabilityn this non-culturaldomainusuallyincorporate he laws of othersciences,suchaschemistry,physics,andgeology.The cultural aspect of formation process concepts has not been appreciablydeveloped.Archaeologistsdo in fact employ interpretiveframeworks hat includeassumptionsabout theseformation processes. These assumptionsare rarely explicit, and therefore do not readilylendthemselves to testing and modification. The small body of explicit concepts deals almostexclusivelywith chronologicalrelationships Dunnell 1970; Rowe 1961, 1962). Whilethe correctchronologicalplacementof past events is necessary or a successfulreconstructionof pastculturalsystems, especiallyasthey change, t can in no way be consideredsufficient.What is advocatedhere, and elsewhere (Binford and Binford 1968:1-3), is a concern withexplaininghow the archaeologicalrecord is producedin termsof explicit models, theories, andlaws of how culturalsystems operate.This paperwill be an attempt to suggest, n the regrettableabsenceof such a corpusof rigorouslytested concepts (Fritz andPlog 1970;Aberle 1970), someways by which we may begin to think about the questionsraisedso far. Hypotheses will bepresented, which, if adequately tested, can contribute to the eventual synthesis and

156

8/22/2019 Schiffer 72

3/11

ARCHAEOLOGICALONTEXTsystematization of an archaeological theory having both explanatory and predictiveutility.Without this kind of a scrutinizableframework,any use of archaeologicaldata to infer pastactivitiesor organization s highly suspect (Binford 1968a) and subjectto interminabledispute.Explicit generationand use of formationprocessconcepts, and other branchesof archaeologicaltheory,will allowgenuinely ntersubjective tatements o be made about the past.

PRELIMINARY EFINITIONSSome preliminaryconsiderationsof a generalnature are requiredat the outset. For presentpurposes,a cultureis viewed as a behavioral ystem of self-regulatingndinterrelatedubsystemswhichprocuresandprocessesmatter,energy,andinformation Miller1965a, 1965b;Clarke1968).A self-regulatingor homeostatic system is defined as one in which at least one variable ismaintainedwithin specifiablevalues despite changes n the system'senvironment Miller1965a;Hagen1961).The values of subsystemvariablesare maintainedwithin their rangesby the performanceof

activities. An activityis a transformation f energy,minimally nvolvingan energysource,oftenhuman,actingon one ormoreproximatematerialelements. An activitymaybe viewedsimplyas apatternedenergy transformation White 1959), which servesto maintainthe values of systemvariables. An activity structure is defined as the activitiesperformedand their performancefrequencies,usuallywith reference o a site, but not of necessity.I define elements to includefoods, fuels,tools, facilities,machines,humanbeings,andallothermaterialswhich one might list in a completeinventoryof a culturalsystem.A provisionaldivisionof elements into the categoriesof durablesand consumableswill be made for lateruse. Durableelements aretools, machinesandfacilities-in short, transformers ndpreservers f energy(Wagner1960). Consumablesare foods, fuels, andothersimilarelementswhose consumptionresults n theliberation of energy. Althoughnumerousotherdimensionscouldbe usedto delineatecategoriesofelements,that is the task of an investigator ttempting o solvemorespecificproblems hanthoseof concern here. It should be mentioned, however, that elements are often compounded intolarger,more complex elements, and that complex elements may be furthercompounded intohierarchies f elementcombinations.In order to continue activity performance, and hence maintain the values of subsystemvariables, t is necessaryto replaceelements which become exhaustedor otherwiseunserviceable.The failure of an element to articulateproperly with other elements is a significantbit ofinformationto the system, which initiates the performance f otheractivitiesresulting ventuallyin element replacement, or activity structurechange. That it also initiates the processes ofdiscardingthe replaced element is a significantbit of information to archaeologists.Whatitintroducesis the life cycle or history of any element-the stages of its "life" within a culturalsystem-and how these relate to the eventual transitionof elementsto the archaeologicalecord.Systemic context labels the condition of an elementwhichis participatingn a behavioralystem.Archaeological ontext describesmaterialswhichhavepassedthrougha culturalsystem,andwhicharenow the objectsof investigationof archaeologists.

THEMODELWhileone may readilyvisualizethe flow of pottery, or food, orevenprojectilepoints, throughaculturalsystem, it is the case that all elements enter a system, are modified,brokendown, orcombinedwith other elements,used, and eventuallydiscarded.Thisis so evenfor thoseelements,such as houses, which at certainpoints in time appear o be permanent eatures.Thisobservationcan provide the basis for the constructionof a simple flow model with which to view the life

historyof anyelement, andaccountbehaviorallyor the productionof the archaeologicalecord.Iacknowledge hat the model to be presentedhere,andthe behavioral omplexitiesto whichit callsattention,has been anticipatedby Lewis R. Binford(1968a:21) andK. C. Chang 1967:106-107);a generaldebt is acknowledged o WalterTaylor's 1948) seminalwork.

Schiffer] 157

8/22/2019 Schiffer 72

4/11

ameRicanantiquityFor analytical purposes, the activities in which a durable element participates during its life, orsystemic context, may be broadly divided into 5 processes: procurement, manufacture, use,maintenance, and discard. A process consists of one or more stages, such as the stages ofmanufactureof a ceramic vessel. A stage consists of one or more activities,which for some

analyses might be further broken down.The model for consumable elements is parallel to and adapted from the model for durableelements. Such an adaptationis necessary to bring the model into congruencewith standardterminology.One would scarcelyrefer, for example, to the manufacture nduse of poachedeggs.The terms for each process of the flow model for consumablesare:procurement,preparation,consumption,and discard. Becauseconsumptionoccursbut once during he systemiccontext of aconsumable, the maintenance process has been deleted. For the sake of convenience, thediscussions to follow will exclusively use the terminology of the durable element model.In addition to the 5 basic processes of systemic context, it will be necessary for some problemsto considerstorageand transport.Storageand transportare activitieswhichprovide, respectively,a temporalor spatial displacementof an element. Transportand storagemay takeplacesinglyorin combination between any 2 processes, stages, or activities of one stage.Not all elements follow a unilinear path through a system. Some are rerouted at strategic pointsto processesor stages throughwhich they have alreadypassed.Archaeologists ncounter tems ofthis sort frequently;this condition is often known as reuse.Two varietiesof reuse,recyclingandlateral cycling will be defined here.Recycling labels the routing of an element at the completion of use to the manufacture processof the same or a different element. In our system, preciousmetals and gems arerecycled.Somesystems recycle potsherds,bifaces, ground stone, and many other elements, most of which arerouted to the manufacturingprocessesof different elements.Sometimesthe use modification ormaintenanceactivitiesof one element can be viewed as the manufacturing ctivitiesof another.Continuedretouchingof a scraperwill result in an implementunsuitedfor furtheruse. But in thisform, the element may be adapted for reuse in some other activity.

SYSTEMIC I ARCHAEOLOGICALCONTEXT CONTEXT

I ,-Irecycling ---

toteralrcycling

---- procurement - *- manufacture - *- use - discard - o- refuse

JI I ,maintenance

LEGEND- System under analysis

Opportunity for storageandor trnsportnd/or transportFig. 1. A flow model for viewing the life cycle of durable elements.

158 [Vol.37, No. 2, 1972

8/22/2019 Schiffer 72

5/11

ARCHAEOLOGICALONTEXT

~~~~~~~~~~I SYSTEMIC I ARCHAEOLOGICALCONTEXT CONTEXT

recycling . -

| laterol IV cyclingI ? t I---- procurement-*- preparation * consumption - "discard-* - refuse

LEGEND- System under analysis

Opportunity for storageand/or transport

Fig. 2. A flow model for viewing the life cycle of consumable elements.Lateral cycling describes the termination of an element's use (use-life) in one set of activitiesand its resumptionin another, often with only maintenance,storage,and transport ntervening.Specifically,reference s made to the movementof clothing, tools, furniture,and otherelements,which in simpleandcomplex systemscirculateamongand betweensocialunits, classes,and castes.Figs. 1 and 2 illustrate the completed basic models for durables and consumables, respectively.I wish to emphasize that these models are only simplifications of a stubbornly complex reality.They are not likely to fit neatly the sequencesof activities in which elements of all culturalsystemsparticipatewithin theirsystemiccontexts.Some regular asesof apparentdivergence romthe models can be listed. For example, an element need not pass throughall processes.Tradeelements are those which have no manufactureprocess in the recipient system. Some elementshave no manufactureprocess in any system;unmodified stone used in construction,and somepercussion lakingtools arecommonexamples.Some items will be discardedwithoutmaintenanceeverhavingbeenperformedon them. Defectiveitemsmaybe discardeddirectlyaftermanufacture.

An element havingno use process,such as anunutilizedflake,is designatedaswaste;this is not toimply that such items aredevoid of information,only that they are an unusedby-productof someactivity.At the termination of an element's use-life (assuming no reuse), elements will be discarded.Refuse labels the post-discard onditionof an element-the condition of no longerparticipatingna behavioralsystem. The normal flow of elements throughthe system in the manneroutlinedabove accounts for most of the materials that become a part of the archaeological record.Although most refuse material consists of those elements which have broken down or worn outduringuse, whole, apparentlyserviceable tems are frequentlyencountered n excavation. Thesematerialspresentadditionalproblemsof explanation.Some mayhave been accidentallydeposited,or theirpresencemay reflect change-an element has become obsolete anddiscarded. n our ownculturalsystem, undamagedand potentially reusable elements whose recyclingcosts arehigherthan replacemenicosts arediscarded."No depositno return"bottles area notoriousexample.Thepresence of such items in the archaeologicalrecord is still accounted for in terms of the flowmodel, but their formalpropertiesmust be accountedfor in termsof economicprinciples.These

159Schiffer]

8/22/2019 Schiffer 72

6/11

american antiquityvariousfactorsare importantand worthy of furtherinvestigation,yet they do not accountfor agreatdeal of the remaining"anomalous"material ound in archaeologicalontext.Elementsdiscardedwith the deceasedafterceremonialuse providea significantsourceof intactelements in archaeological context, especially among simple systems. The subject of graveaccompanimentsand their relationshipsto other aspects of the system that discardedthem,especially social organization,will not be discussedhere;althougha comprehensive reatment slong overdue.The principal set of variables responsible for the presence of usable elements in thearchaeological record is obviously that having to do with the abandonment of the site.Archaeologicalcontext includes all the materials found in a site, whether or not they are inspecializeddiscard ocations and whetheror notthey have beendeliberatelydiscardedby the pastoccupants of a site. It is well known, for example, that elements are found in every stage ofmanufactureand use. Theway in whicha site is abandoned-the variablesoperativeat the time theoccupantsleave the site, or die without replacement-hasdemonstrable ffects on the kinds andquantities of non-discardedelements found in archaeologicalcontext. Elements which reacharchaeologicalcontext without the performanceof discard activities will be termed de factorefuse.Robert Ascher'sprovocativeSeristudy(1968) suggestsa hypothesis,which, whengeneralizedsrelevanthere: differentialabandonmentof a site changesthe normalratiosof elementsin variousprocessesof their systemic contexts and the normalspatialdistributionof elements.Specifically,reference is being made to activities which result in the removalof raw materialsand usableelementsfrom abandonedareasof the site andtheirreusein the still-occupiedportion.At the timeof their abandonment, these elements were still in systemic context. We would expect to findrelativelyfewerelementsin pre-discardprocessesof systemiccontext, that is, lessde facto refuse,in sites which undergodifferentialabandonment.On the otherhand,sites abandonedrapidlyandcompletely as a result of some catastrophewill have relativelygreaternumbersof elements inmanufacture,use, and maintenanceprocesses.Pompeii comes to mind as anexampleof thiskindof abandonment, in which no change occurred in the provenience of elements or theirdistributionsamongthe varioussystemicprocesses.More often, abandonment,even if sudden, involvesthe removalof some elements and theirtransportto another site or sites. The kinds and quantitiesof elements so removed should berelatedsystematically o othervariablesoperativeat the timeof abandonment.Amongthemmightbe expected: distance to the next site, season of movement, size of emigrating population,technological developmentof both donor and recipientsites, means of available ransportation,and other variables.Morecomplex models will have to be devisedto account for the effects thatprocessesof abandonmenthaveon the formationof the archaeologicalecord.

SPATIAL MPLICATIONSPerhaps the most important aspect of the notion of systemic context is that there is aspecifiablespatial ocation, or locations,for eachprocessthroughwhich anelementpasses.The term "location"is used here in its broadestpossiblemeaning.A locationcanbe a point ona site, or it canbe a set of points.It can alsobe anentiresite if, duringone process,anelementhasthe same chance of being found at one place on the site as any other. Such a concept of spatiallocation mightbe better expressedas a set of probabilities or findingan element or classof likeelements at any point on a site surfaceduringa particularprocessor stage.A site or otherunit ofspatialanalysis s dividedinto squaresof equal areaandtheprobabilityvaluesfor eachsquareareindicated.The smallerthe squares,the greaterthe potentialaccuracy see ColeandKing 1968 fornumerousexamplesof how one mightmodel distributions).This flexibilityin depicting ocations

allows the frame of referenceto be shifted conveniently to any variableof interest to suit theneeds of the investigator; he metate of a woman, or the metates of a village.The relationshipsbetween the locations of each process or stage for an element are complex but I predict thateventuallydeterminaterelationshipsbetweentheseandbehavioral ariableswill be specifiable.

160 [Vol. 37, No. 2, 1972

8/22/2019 Schiffer 72

7/11

ARCHAEOLOGICALONTEXTThe overlappingof these locations for different elements, and the activitiesinto which theyarticulate,reflect a behavioralmatrix of bewilderingcomplexity, even for simplesystems.Whilethis complexity presentsproblemsfor some uses of archaeological ata,it alsoprovidesa hithertoneglectedsourceof information or generatingandtestingbehavioralhypotheses.Archaeologistsare often able to reconstruct he manufacturing ctivitiesof elements when theyare recovered n v-arioustagesof manufacture,often in associationwith wastematerials.That thevariousstagesand processesof an element'ssystemiccontext should be reflectedspatiallyaswellhas been less frequentlyused as a basis for generatingor testinghypotheses.A projectile pointfound in associationwith an antler flaker and tiny pressureflakes of the same materialin ahabitationstructure s a differentprojectile point from one found in a midden,or another ounderodingfrom the wall of an arroyowith no other associatedculturalmaterial.In the firstcase,onewould be dealingwith the location of some manufacturingactivities,the second a location ofdiscardactivities,while in the thirdperhapsthe location of use. The cultural nferencespossiblefrom a projectilepoint in each processaredifferent,asare the potentialhypotheses againstwhichthesemorphologically imilarelementsmaybe broughtto bear asevidence.In anotherexampleof the relationshipbetweensystemiccontext andthe spatialdifferentiationof culturalbehavior,one may take the case of the simplesubterraneantoragepit. Becausesuchpits occupy the same location duringall processes,one is permittedto statewith certaintythatapit was dug, used, andrepairedby the inhabitantsof the site at the sameproveniencewhereit wasfound in archaeologicalcontext. This is hardly a revelation,yet I suggestthat such rigorousjustificationfor a primarybehavioralnference s necessaryf we are to succeed n formulatingandanswering he kinds of questionswhich are now being asked about past culturalsystems. Thearchaeologicalrecord will produce sets of information on subjects which we can scarcelycontemplate at the present, when, and as, progress is made in buildingmodels to relate theproductionof the archaeologicalrecord to culturalbehaviorin the past-models which includeexplicitreference o the spatialdimensionof culturalbehavior.

REFUSEDISPOSALPATTERNSIf this model of culturalelement flow and its spatialaspect areto be of value in enablingus togain knowledge of the past, they must illuminatesome of the questionsposed earlier.Thesequestionsand others will be examined andrephrasedn termsmoreamenable o treatmentby theconcepts sketchedout here. Additionalhypotheses will be introducedas the need for themarises;they areintended,asare allhypotheses,to be suggestiveand not definitive.Weshall now return to the questionraisedoriginally n the quotationfromBinford; hat is, towhat extent can remainsat a site be expected to occur at theiruse locations as opposedto anyother, when found in archaeological ontext? We aim to know whatsome of the determinants fvariabilityin patternsof refuse transportand disposal are. I shall distinguishbetween primaryrefuse and secondary refuse. Both refer to elements whichhave been discarded compareto defacto refuse)but, in the case of secondaryrefuse, the location of final discard s not the sameasthe location of use. Primary efuse s materialdiscardedat its location of use (Fig. 3).I believethat the generalproblemof refusedisposalmay be seenas the balancingof 2 majorsetsof variables.The particular olutions arrivedat by site occupantsfor handling he by-productsofactivity performancewill takeinto consideration he ease of movingthe activityor activitiesversusthe easeof movingthe refuse.Let us assume that there is a site at which only one activity is beingconductedby a singlepersonduringbriefperiodsof the year. In this case one might expect few pressures o favorthedevelopmentof a separate location for the final discardof elements replaced duringactivityperformance.One would obtain a generalcorrespondencebetween the location of use and thelocation of final discardfor elements used in that activity.Let us, then,increase he populationof

the site to a smallvillage,and increasethe intensityof occupationto year-round.n this instance,one would expect such factorsas the necessity for unrestrictedaccessbetweenprincipalactivityareas,sanitation,and competitionfor scarceactivity spaceto placea premiumupon the transportof at least some of the materialsand their discardat anotherlocation. Moderncities providethe

Schiffer] 161

8/22/2019 Schiffer 72

8/11

ameRitcanantiquity

I SYSTEMIC I ARCHAEOLOGICALCONTEXT CONTEXT

I discord I primary refuse-- procurement - *-manufacture -e *e- use|J I\ \. \ tronsport - discard-- secondary refuse

t defacto refuse

LEGEND-- System under analysis

Opportunity for storageand/or transport

Fig. 3. Simplified flow model for explicating the differences between primary, secondary, and de facto refuse.

extreme example, as we know it today, where almost no elementsarediscardedat theirplacesofuse within the site; consequently, almost all archaeological context material is secondary refuse.The general principle which these hypothetical cases illustrate is that with increasing sitepopulation (or perhaps site size) and increasing intensity of occupation, there will be a decreasingcorrespondence between the use and discard locations for all elements used in activities anddiscarded at a site. In addition, there will be an increasing development of specialized discard areas,occupations,and transportnetworks.Fromthisprinciple,admittedlyunpolishedanduntested,wecan predict that limited activitylocations(Wilmsen1970) suchaskill andbutchering ites,quarrysites, and many seasonally occupied sites, will consist largely of primary refuse. A majorcharacteristic f these sites will be the repeatedclusteringsof elements n discreteandoverlappinglocations.Let it be given that many sites of many systems had at least moderately developedrefusetransportand disposalactivitiesand, as a result,elements used in many activities were removedfrom their locations of use. The questionthat would appearto be before someone interested ininferring he past activitystructureof such a site is, To what extent are elementsassociated n usealso associated as secondary refuse? No definitive answersare presently at hand, though onehypothesis may accountfor some elementassociationsn secondaryrefuse.If storageintervenesbetween the terminationof anelement'suse-lifeand its finaldiscard, hereis a likelihood that one ormore other elements of the sameactivitywouldhavebeen replacedandstoredalongwith the first element awaitingfinal discard.Therefore,as there is a decrease n theratio of final discard requency to replacement requency of one or moreelementsof anactivity,there is an increasingprobability that several elements, especially those with short use-lifeexpectancies,will be discardedat the same time and in the sameplace withinsecondaryrefuseareas. The optimum conditions for element association as secondary refuse occur in modernindustrial ocieties wheremanystorageandtransport tepsintervenebetween elementreplacementand final discard.Most other activities in most other culturalsystems will result in secondaryrefuse which lies somewherealong this continuumof activity-based lement association.Future

162 [Vol. 37, No. 2, 1972

8/22/2019 Schiffer 72

9/11

ARCHAEOLOGICALONTEXTresearch in both extant and extinct cultural systems must be undertaken to provide moreknowledgeon the regularities f dumpingbehavior.The relative frequenciesof elements, or element fragments,found as primaryor secondaryrefuse,are rawdata for manystatementsmadeabout the past.I thinkit is fair to questionanyuseof this informationuntil we know the ways in whichrefuseelementfrequenciesreflectthe systemof which they were once a part. A general olution to thisproblembasedon previoushypotheses,whichwill admitof manysourcesof exceptions,cannow be presented.Assuming hat there is no changein the activitystructureduringthe occupation of a site, andthat there is only one refusearea,which maybe the entiresite, the ratiosof elements in that areawill correspondto theirrelativereplacement requencies.For example, althoughonly 1 mano isused with 1 metate atanygiventime, the ratioof discardedmanos to discardedmetates(assumingno recycling)will correspond o how often one is worn out andreplacedwith respectto the other,which may be 6 or 8 manospermetate.Thismodel is complicatedby elementswhich haveseveraldiscard locations, one or more of which are not known or accessible to the investigator.Particularlyacute is the problem posed by projectilepoint disposal patterns. Any statement,whetherfor chronological ontrol,culturalaffiliation,activityreconstruction,or the measurementof a past systemicvariable,requiresstrict considerationof themultiplediscardareas or thiskindof an element. A potentially fruitful topic for investigationis the conditions under whichprojectile points, or anysimilarelement,will be discarded t a habitationsite. It mayturnout thatsuch points are a perfectlyrepresentative ampleof allpoints in use;yet at the moment,we reallydo not know one way or the other.In offering inferences about past activity structure,reportshave sometimesbeen made thatritualactivities were infrequentor absent.A different nterpretations possible.I hypothesizethatdurable elements used largely in ritual activities will have on the averagea longer use-lifeexpectance than non-ritualdurableelements of the same system. If this is the case, then even ifritualactivities were presentand frequent,non-ritual lementswouldbe expectedto predominatedisproportionatelyas refuse, simply as a result of differentialreplacement frequencies. Anystatement assertingthe absence or infrequent performanceof any activity should be treatedskepticallyuntil the biases ntroducedby differentialreplacement requenciesandmultiplediscardlocations havebeen takeninto account.

CONCLUSIONArchaeologistshave gone from the one extreme of viewinga site as spatiallyandbehaviorallyundifferentiatedrubbish to the other extreme of viewing remainsas mostly reflecting theirlocationsof use in past activities. At this point, it appears hat neitherextreme is often the actualcase. Clearly, though all remains in a site are refusewhen uncovered n archaeologicalontext,when viewed by the model and hypotheses presented here (and used implicitly by manyinvestigators) hey are potentiallymuch more. In orderto realize this potential,we will have tolink archaeologicalontext material o behavioral ndorganizational ypothesesaboutelements nsystemiccontext.I submit that this linkingprocess s the centralproblemof archaeologicalnference see Binford1968b for similarstatements). Once high probability statements about activity structuresarepossessed,hypotheses regarding he compositionof task groups,theirmeans of recruitment,andhow they are structured within the total system organization, and especially how theseorganizations hange,will be capableof preciseformulationandarchaeologicalest.Whetheronebegins at the level of archaeological ontext material,or with models of system organization ndchange, the form of the final inference or tested model will be similar: statements aboutpastorganizationor other systemic properties arelinked by argumentsof relevance(Binford 1968b;Fritz 1968; Schiffer 1970) to the activitystructure.The activitystructure,n turn,is linked to thearchaeological ontext databy formationprocessconcepts.The constructionanduse of formationprocessconceptsalongthe linessketched out abovewillallow the rigorousjustification of our inferences.Without a base of explicit, logically relatedcredible laws about the formation processes of the archaeologicalrecord, debates about the

Schiffer] 163

8/22/2019 Schiffer 72

10/11

ameillcanantiquity [Vol. 37, No. 2, 1972validityof an inference,or any use of the datafromthe record,canonly focus on epiphenomenaor ad hominem arguments Binford 1968a). I would hope that the crude first approximations oexplicit formationprocess concepts presentedhere will stimulate a round of vigorouscriticismaimed at improvingthe conceptual tools with which we manipulatethe remainsof past culturalsystems. As more sophisticatedand comprehensivemodels are developed, confidence will begained in the uses to which we put the data of the archaeological record.

Acknowledgments. I am indebted to the following individuals for their patient and instructive criticisms ofvarious versions of this paper: Daniel C. Bowman, David A. Gregory, John A. Hanson, William A. Longacre,Henri A. Luebbermann, Paul S. Martin, David R. Wilcox, and the Editor. I am solely responsible for themisconceptions which persist. In the context of solving a different problem, Michael B. Collins (1971) hasindependently developed a lithic-specific flow model similar to the general model presented here; I have profitedfrom discussions with him. Sharon Urban kindly prepared the figures. I gratefully acknowledge the support ofthe Field Museum of Natural History and the National Science Foundation (Grant GY-7225). I am especiallyindebted to the director of the Field Museum Southwest Expedition, Paul S. Martin, for his constantencouragement and for providing an unmatched climate of intellectual freedom.

Aberle, David F.1970 Comments. In Reconstructing prehistoric pueblo societies, edited by W. A. Longacre, pp. 214-223.University of New Mexico Press, Albuquerque.Ascher, Robert1968 Time's arrow and the archaeology of a contemporary community. In Settlement archaeology, editedby K.C. Chang, pp. 43-52. National Press Books, Palo Alto.Binford, Lewis R.1962 Archaeology as anthropology. American Antiquity 28:217-225.1964 A consideration of archaeological research design. AmericanAntiquity 29:425-441.1968a Archeological perspectives. In New perspectives in archeology, edited by Sally R. and Lewis R.Binford, pp. 5-32. Aldine, Chicago.1968b Some comments on historical versus processual archaeology. Southwestern Journal of Anthropology24:267-275.Binford, Sally R., and Lewis R. Binford1968 Archeological theory and method. In New perspectives in archeology, edited by Sally R. and Lewis R.Binford, pp. 1-3. Aldine, Chicago.Brose, David S.1970 The Summer Island site: a study of prehistoric cultural ecology and social organization in the northernLake Michiganarea. Case WesternReserve University Studies in Anthropology 1.Chang, K.C.1967 Rethinking archaeology. Random House, New York.Clarke,David1968 Analytical archeology. Methuen, London.Cole, John P., and Cuchlaine A.M. King1968 Quantitative geography: techniques and theories ineography. John Wiley & Sons, New York.Collins, Michael B.1971 The role of lithic analysis in socio-cultural inference. Paper presented at the 1971 meeting of TheSociety for American Archaeology, Norman.Dunnell, Robert C.1970 Seriation method and its evaluation. American Antiquity 35:305-319.Fritz, John M.1968 Archaeological epistemology: two views. Unpublished M.A. thesis. Department of Anthropology,University of Chicago.Fritz, John M., and Fred T. Plog1970 The nature of archaeological explanation. American Antiquity 35:405-412.Hagen, Everett E.1961 Analytical models in the study of social systems. The American Journal of Sociology 67:144-151.Hill, James N.1970a Prehistoric social organization in the American Southwest: theory and method. In Reconstructingprehistoric pueblo societies, edited by W.A. Longacre, pp. 11-58. University of New Mexico Press,

Albuquerque.1970b Broken K Pueblo: prehistoric social organization in the American Southwest. University of Arizona,Anthropological Papers 18.Hole, Frank, and Robert F. Heizer1969 An introduction to prehistoric archeology. Holt, Rinehart and Winston, New York.Longacre, WilliamA.1970 Archaeology as anthropology: a case study. University of Arizona, Anthropological Papers 17.

164

8/22/2019 Schiffer 72

11/11

Schiffer] ARCHAEOLOGICAL CONTEXT 165

McPherron,Alan1967 The Juntunen site and the Late Woodland prehistory of the Upper Great Lakes area. University ofMichigan,Museum of Anthropology, Anthropological Papers 30.Miller,James G.1965a Living systems: basic concepts. Behavioral Science 10:193-237.1965b Living systems: structure and process. Behavioral Science 10:337-379.Rowe, John H.1961 Stratigraphyand seriation. American Antiquity 26:324-330.1962 Worsaae's aw and the use of gravelots for archaeological dating. American Antiquity 28:129-137.Schiffer, Michael B.1970 Cultural laws and the reconstruction of past lifeways. Paper presented at the 1970 meeting of TheSociety for American Archaeology, Mexico City.Struever, Stuart1968 Woodland subsistence-settlement systems in the Lower Illinois Valley. In New perspectives inarcheology, edited by Sally R. and Lewis R. Binford, pp. 285-312. Aldine, Chicago.Taylor, WalterW.1948 A study of archeology. AmericanAnthropological Association, Memoir 69.Wagner,Philip L.1960 The human use of the Earth. The Free Press, Glencoe.

White, Leslie A.1959 The evolution of culture. McGraw-Hill,New York.Wilmsen,Edwin N.1970 Lithic analysis and cultural inference: a Paleo-Indian case. University of Arizona, AnthropologicalPapers 16.