Pastoors Weniger 2011 JAR

CaveArtinContext:MethodsfortheAnalysisoftheSpatialOrganizationofCaveSitesAndreasPastoorsGerd-ChristianWenigerSpringerScience+BusinessMedia,LLC2011Abstract Investigations of prehistoric cave art have long neglected the sur-rounding context: space, archaeological objects, and imprints. As a result, anintegrativestructural approachthat analyzes caveart aspart of ananthropomor-phizedlandscapehasnotbeenavailable.Thisarticledrawsonurbanplanningandthe physiology of the human eye to provide an innovative archaeospatial analysis ofcavesites. Aset ofrelevant features fromthecavesofBedeilhac, Fontanet, andLePortelwasselectedanddened(lightzone, chambertype, pathnetwork, modeofmovement, and available space). An analysis of the prehistoric remains in the cavesallowsthereconstructionofdifferentconcentrationsofhumanactivities(caveart,archaeological objects, andimprints). Theprojectionoftheseconcentrationsontothestructuredmapofthecavesresultsinfourtypesoflocations:drawinglocation,supplylocation, drawinglocationwithsubstantial activities, anddrawinglocationwithconsumptionactivities. This approachopens newavenues for thearchaeo-logical perceptionof cavesandtheir inhabitants: Upper Paleolithichumanswerevery familiar with caves and probably followed a master plan during their stay in thedark.Keywords Prehistoriccaveart Spatialorganization Landscapearchaeology UseofcavesIntroductionCaves are [] natural cavities in the earths crust that are to a large extentenclosed by compact masses of stone and rock; they are lled with eitherair, sediment, or water [] andare large enoughtopermit entrybyhumansA. Pastoors(&) G.-C.WenigerStiftungNeanderthalMuseum,Talstrasse300,40822Mettmann,Germanye-mail:[email protected] 3JArchaeolResDOI10.1007/s10814-011-9050-5(Wrede 1996, translationfromthe GermanbyK. Schneider). Nevertheless, theworldof thedeepest cavesisaforeignenvironment for humans; thesolar cycle,which is such a structuring feature of life above ground, is completely absent; thereis no night and day. Furthermore, in contrast to the outside world, the world of cavesis constant. Slight changes in daily temperature and air humidity are barelyperceptible,andconstantdarknesspersistsintheaphoticpartsofthecave.Indeed,regardless ofthe time spentinthisblackness, thehumaneyeneveradjusts. In spiteof this hostile environment, andpossiblyexactlyfor that reason, humans havealways been attracted by caves. In the past, caves have provided, and still provide, aplaceofconcealment, accommodation, burial, shelter, andaplaceforceremonialgatheringsandinitiations;theyhavebeenusedasworkshops,sportsfacilities,andmore(BonsallandTolan-Smith1997;Pasda2004).This article explores the relationship between humans and caves during theUpper Paleolithic by using an integrative approach. The aimis to reveal howhumans adaptedtothe alienenvironment of deepcaves andhow, via intersitecomparison, isolated phenomena and repetitive patterns of behavior can beidentied. ThreecavesitesintheFrenchPyreneesBedeilhac, Fontanet, andLePortel, allof whichhave yieldedcultural remains fromthe Magdalenianformthearchaeologicalbasisforthestudy.Art, aspartof cultural expression, is rst visible in archaeologicalrecordsintheAurignacian. Eventoday, inspiteof thelong-standingresearchtraditioninthiseld, the meaning of these representations and the interpretation of the surroundingcontext are still extremely speculative and inuenced by the intuition of theresearcher. This decit is due not least to the prevalent approaches, wherebyindividualguresarerstdescribedingreatdetailandsubsequentlyinterpretedonthebasisofhighlypersonallevelsofexperience.Todate,researchstillfocusesonmeticulousrecordingofguresandtheirmodesofexecution.Majorenhancementsintechnologyhaveevenledtotheatomizationof theguresintotheir smallestgraphical units (e.g., Fritz and Tosello 2007). The implementation of these graphicalexpressions into a wider frame of human behavior in caves is still pending, althoughthesignicanceofcavesasspaceswithfrequenthumanactivitiesandcavearthasbeen stressed by several Paleolithic researchers (e.g., Bahn 2003; Lorblanchet1995). Research needs an integrative approach, linking art and other forms ofhumanactivitiesembeddedinthenaturalspaceoftheentirecave.Sincethecompilationofthe apports mobiliersfromtheVolpcavesbyBegouenandClottes (1981), wehavebeenawarethat amultitudeof different prehistoricremains can occur in the context of cave art, even though they are rarely preserved.Thisisalsoduetothefactthat,inmanycases,researchershaveratedthevalueofthe artwork itself considerably higher than objects and traces of daily life (Begouen1926). Indeed, inthe past thisvery approach has served as justication for blowingscientic caution to the wind and accepting the destruction of an apparently inferiorsource of information in favor of investigating marvelous artistic expressions.Unfortunately,this behaviorcanstillbeobservedattimes.GrotteChauvet(Clottes2001; Geneste2005), TucdAudoubert (Begouenet al. 2007, 2009), Les Trois-Fre`res(BegouenandClottes1981),Bedeilhac(Malvesin-Fabreetal.1953;Sauvet2007), Fontanet (Delteilet al. 1972; Vialou1986), LePortel (Beltranet al. 1967;JArchaeolRes1 3Vezian1955), LaGarma(AriasCabal1999;AriasCabaletal. 2003), andCussac(Aujoulat2002)allshowthatthequantityandqualityofanthropogenictracescanbe quite exceptional. It is quite possible that some of these traces can provide muchgreatercontributionstotheinterpretationofcaveartthantheguresthemselves.AcloserlookatthecavesofBedeilhac, Fontanet, andLePortelprovidesgoodexamples. The descriptions of the caves themselves are intuitive and vague, and thesame chamber is often described differently by different authors. The GalerieprincipaleinBedeilhac,forexample,isdescribedas vaste(Ministe`redelaCulture1984), proportionsgigantesques(Sauvet 2007), or large, haute(Vialou1986). InBedeilhac, the lateral galleries are describedas dimensions parfois fort exigues(Ministe`redelaCulture1984),reduitsetetroits(Vialou1986),ordacce`sdifcileor defendue par un long passage rampant (Sauvet 2007). Other informationconcerning the cave morphology is given as a kind of circuit promenade through thecave. The information is more literal than structured. Except for cave art, theprehistoric remains are reported on the same level. There is no systematicdifferentiation between the activities. Classications such as aires des sejours,simplefrequentation, sejours prolonges(Begouenetal.2009;Sauvet2007),andIlfallait bien nourrir les artistes (Begouen et al. 2009) are terms that are usedintuitively.Fundamental studiesthat focusedonthestructuringofcavesitesbasedonthedistributionof prehistoriccaveart werepublishedbyLeroi-Gourhan(1965) andRouzaud(1977,1996, 1997). Leroi-Gourhan(1965)suggestedthataclassicationsystemcouldbeachievedbysystematicdataacquisition. Indoingso, headheredto the basic principle that the application of gures provides a cave with astructure. Thus, in the eyes of Leroi-Gourhan, the cave is an ensemble ofchambers comprising entrance zone, in-between areas, central and side areas,passages, andendzone. Ontheother hand, Rouzaud(1996, 1997) analyzedthetraces of prehistoric humans in caves from the perspective of paleospeleology. Thisled himto his denition of chamber types based on lighting and modes oforientation. Hisintegralfeatureischambersize,whichcombinesaccessibilityandilluminationusinganarticiallightsource.Sincethebeginningofcaveart research, thecloserelationshipbetweengureandrocksurfacehas beenoutlined(e.g., Bahn2003; Lorblanchet 1995); Leroi-Gourhan (1971) created the termof la caverne participante to describe thisphenomenon. This alsoapplies tothepositioningof thegureintheimmediatesurroundingspace. Averysophisticatedmethodologywas recentlypublishedbyVilleneuveandHayden(2007). Theyproposethat thevisibilityof singleguresshouldbeanalyzedindetail todrawconclusionsconcerningtheirsocial function,thusdifferentiatingbetweenintimateandpublicpresentation(e.g., Bahn2003). Aquantitative approach was developed by Lorblanchet (2001). He analyzed thedistribution of gures in the Pergouset cave and, in this same context, calculated theabsolutevolume(incubicmeters) of singlechamber units. This resultedintheidentication ofa unit ofmeasurement (number of graphicalunits percubicmeter)thatmakesitpossibletodescribevaryingdensitiesofcaveart.Theneedforwell-denedfeaturestostructurecavesiteswithcaveartcanbenotedinallofthecitedJArchaeolRes1 3authors. Indeed, this ever-growing demand has been conrmed recently byinvestigationsattheGrotteChauvet(LeGuillou2005).BasicprinciplesofaspatialorganizationanalysisofcavesitesThereviewof researchhistoryclearlyshows theneedfor anintegrativespatialapproach for the study of caves and associated prehistoric remains. In addition, thereis a general consensus that the distribution of cave art is inuenced by the prevailingcave topography (Sieveking 1997). Nevertheless, the possibility of formallyinvestigatingthisinuenceisatthesametimeregardedasslight,notleastbecauseof the high natural variability of the caves themselves (Sieveking 1997). Thatsaid, theidenticationofstructuralsimilaritiesshouldmerelybeaquestionofthecorrect angleof visionandviewingdistance. Indeed, variousdisciplinessuchasarchitecture, ethnology, speleology, religious studies, social science, and urbanplanning are all engaged in the study of space. At a workshop dedicated to this topicin 2002, representatives fromvarious disciplines came together (Pastoors andWeniger 2003). There was consensus among participants that space used by humansserves two main purposes: passage and stopover. This duality is a basic principle ofthe use of space andalsopertains tocaves. Inanalyzingcaves, three differentaspects are of particular signicance: the natural structure of the cave, the ability ofthehumaneyetoreact todarkness andarticial light, andtheclassicationofdetectableprehistoricremains.Currently, cave research does not have recognized standards for classifying cavestructure that are oriented toward human needs. The situation is quite different in thecaseofurbanplanning,whereplansaredevelopedanddecisionsaremadeonhowavailableareas canbest beput tousefor humans (e.g., Streich2005). For thisparticulartask, thereisanarrayofmethodological instruments, comprisingmaps,plans, andappropriate terminology. Some components of these instruments areapplicable to the spatial organization analysis of caves. Of particular signicance inurban planning are lines of communication, e.g., pedestrian and cycle paths,railways, androads, that is, elementsofasuperordinatetransit infrastructurethatonlyintheir totalitymaketransportationpossible. This systemincludes featuressuchaslinksandconnectionpoints(crossroads), modesofregulationandcontrol(trafc signs, signals), signposts, and parking and meeting points. In short,communication plays a very special role in all aspects of urban planning. Therefore,althoughtheentirerepertoireof instrumentscannot bedirectlytransferredtothespatial analysis of caves, it serves tohighlight interestingperspectives that mayassistinorganizingthepathnetworkincavesaccordingtolinesofcommunicationandconnectionpoints.Humans are among those beings that must leave the cave on a frequent basis bothto gather food and for temporal orientation, i.e., to ascertain whether it is morning orafternoon, day or night, summer or winter. In the course of human evolution,humanshaveneveradaptedtotheconditionsthat prevail inthosepartsofacavewithout any source of light. For temporal and spatial orientation, the eye is the mostimportant human organ; between 80 and 90% of information about the environmentJArchaeolRes1 3is taken in visually (Griefahn 1996). This also applies to stopovers in a cave,although here all sense organs react with hyper-receptivity to the exceptionalconditions:darknessandsilence.The retinaof the eye is covered by some 6 to 7 million conesthat serve daytimeandcolorvision,andbysome130millionrodsthatfunctionindimlight.Throughadjustment, the human eye can adaptto various states of illuminance, ranging from0.01lx(starlitnight)to100,000lx(fullsunlight)(lx=lux/illuminance).Whereasadjusting to high intensities of light occurs more or less immediately (between 1 and1.5seconds), adaptingtodarknesstakesconsiderablylonger. Althoughadaptationbegins quite rapidly, between 20 and 40 minutes are required to complete the entireprocess. Impaired vision, for example, due to insufcient lighting, can lead topremature tiredness andconsequentlytoa reductioninthe abilitytoreact andconcentrate. Because of dilation(relaxation) of the pupil intwilight conditions,focus depth is reduced, resulting in relative shortsightedness to about 0.5-1.5diopters(BartenbachandWitting2009; Griefahn1996). Thefact that all catsatnightaregrayiscertainlynotduetothecats,butrathertotheretinaofthehumaneye. Inpoor light, onlytherods that functioninthoseconditions react totheirsensors, andtheycanonlyregister graytones. The cones responsible for colorvision remain inactive. For the very same reason, color vision is considerablyimpairedincaves. Furtherimportant factorsthat inuencetheperceptionofcolorarethecolorandbrightnessofthesurroundings.Visible light comprises electromagnetic waves in the region of 380Nm for blue(short waves) to 780Nmfor red (long waves) (Nm=Newton meter). Withincreasingdarkness, rst the short waves (blue) disappear until nallythe longwaves (red) arenolonger distinguishable. For thisreason, redis thecolor mostvisible in poor light (Harten 1997). The threshold of color visibility is given as 3lx.Blue-grayandbrowntones arethenatural fundamental colors of limestoneinadripstonecave,withcalcareoussinterandmineraldepositsaddingtotherepertoireof observable colors. In this environment, attention can best be aroused by the colorred. Inpartsofacavevoidofnatural light, visual perceptionistriggeredlessbycolorthan by thesharpcontrast between illuminatedandnonilluminatedareas,i.e.,light andshade. For humansight, thecognitionof secondarylight sources is ofparticular signicance (Harten1997). Inthe cave, the wall andceilingbecomereectors.DuringthePaleolithic,theprimarysourceoflightcouldonlyhavebeenthe naked ame; torches, tallow lamps, and replaces have all been conrmed fromarchaeological investigations(deBeaune1987; Harten1997). Theluminousuxemittedbythenakedameofacandleliesintheregionof5-15lm(lm=lumen)(Harten1997), andexperiments haveshownthat Paleolithictallowlamps wouldhave reached similar values (de Beaune 1987). Therefore, adhering to theaforementionedthresholdofcolorperception(3lx)andtheknownluminousuxof acandle(5-15lm), it canbededucedat what distancethecolor redbecomesvisibletothehumaneye. Thisisgivenbythedenitionofilluminance(lux, lx):lx=lm/m2. Lux is the illuminance that is generated in an area of one square meterwhenilluminatedbyoneluminousux(1lm). Illuminanceissubject todistanceand decreases quadratically with increasing distance from the light source.Accordingly, inthelight of asinglecandleinanaphoticenvironment, thecolorJArchaeolRes1 3red rst becomes visible at a distance of2.24m [= H(15lm/3lx)]andat the latestat a distance of 1.29m [= H (5lm/3lx)]. The distance at which black (coloring) orengravings are visible cannot be measured. Here, the contrast with surroundings andthe size of the gure are both key factors. Nevertheless, it is remarkable that a nakedame of 5-15lmis sufcient to generate an illuminance of 0.01lx that isobservable at a distance of between38.73m[= H(15lm/0.01lx)] and22.36m[= H5lm/0.01lx)]. Thisvaluecorrespondstothedarknessofanight illuminatedsolelyby starlight(Griefahn 1996) andisadequate forsimpleorientationpurposes.However, considering the often-difcult conditions encountered ina cave, it isunlikely that this illuminance would be sufcient to allow for controlled movement.Due to the often-difcult topography of a cave, it is by no means easy to pinpointexactly the amount of illuminance essential for controlled movement. It is assumedthatanightilluminatedbyafullmoonwithavalueof0.25lx(Griefahn1996)issufcient forthispurpose. Candlelight offersanilluminanceofthat valueuptoadistanceof between4.47and7.75m. Thesecalculateddistancesassumethat thearea upon which the light falls provides optimal reection (100% reectivity); this iscertainly not the case for the gray-brown walls of caves. For that reason, we suggestthatthisdistancebebroughtdownto4m.Cave sites not only contain rock art but, in favorable conditions, also haveyieldedothertypesofremainsleft behindbytheirprehistoricvisitors(e.g., AriasCabal 1999; AriasCabal et al. 2003; BegouenandClottes1981; Begouenet al.2009).Thesecould be usedas a frameofreferencefor artwork. Inadditionto caveart, twogroupsofprehistoricremainscanbedistinguished:archaeologicalobjectsandimprints. Theformer arepart ofthematerial culture(physical residue) (e.g.,Eggert 2001, p. 52), though Begouen and Clottes extend this term to include apportsmobiliers, by which they refer to intentionally and unintentionally left marks(Begouen and Clottes 1981, p. 157). These include imprints of the peoplethemselves. Animprintisgeneratedbytheaccidental ordeliberatepressureofanobject or bodypart against a plastic surface. The imprints of bothobjects andhumans are of considerable signicance for archaeological research. One of the bestknownplaces tohaveyieldedthis typeof evidenceis theupper galleryat TucdAudoubert, whereaverydiversespectrumof imprints is preserved, includingremoved or moved bones and stalagmites, projectiles, feet, heels, toes, ngers(includingangernail), andknees(Begouenet al. 2009). Withduerespect totheveryparticularcircumstancescontributingtopreservationofthistypeofevidence,inthefollowing, imprints aredesignatedas anindependent groupof prehistoricremains.Ideally, all three categoriescave art, archaeological objects, and imprintscancontribute to an analysis of the spatial organization of a cave. These differentcategories can vary infrequency. The ne-grained breakdown into quantitativeclasses based on statistical calculations proposed by Lenssen-Erz (2001) in hisanalysis of the prehistoric artwork at the Brandberg in Namibia seems to be the mostsuitable methodfor asensiblearrangement of rockart intoquantitativegroups.Unfortunately, thiswasnot possibleforthecavesat Bedeilhac, Fontanet, andLePortel, wherewehavetomakedowithcoarse-graineddescriptions fromearlierpublications.JArchaeolRes1 3Concentrationsofcaveart,archaeologicalobjects,andimprints canbe analyzedbothquantitativelyandqualitatively.Aconcentrationisdenedbythepresenceofmore than one prehistoric remain per unit area. The dimension of the unit area has tobedeterminedwithrespect totheindividual situation. Assuch, activitiesat eachconcentrationcanbereconstructedbytheanalysis of operational processes. Foropen-air sites, Binford (1983) proposed differentiating between concentrations withsubstantial activities and those with qualied activities. Whereas complete reductionsequences of lithicblankproduction, theproductionof retouchedtools, andtheoccurrenceofpredominantlycompleteskeletonsofpreyarecharacteristicfeaturesof concentrations with substantial activities, isolated skeletal elements andfragmentary reduction sequences of lithic production are classied as qualiedactivities. In addition to these substantial activities, two classes of qualiedactivities(consumptionanddrawing)havebeenidentiedinthecavesunderstudyhere. Whereasconsumptionactivityinvolvestheutilizationofprovisionsbroughtintothe caves without important blankproduction, the termdrawingactivitysubsumes thecreationof gures or other artisticrepresentations onthewalls oroors of the caves. Due to the large repertoire of methods used in their completion,as well as thevast arrayof diversemotifs found, thesefeatures providefurthermeansforpatterningaccordingtovisibilityanddimensionsofmotif(Bahn2003;VilleneuveandHayden2007).AtoolkitfortheanalysisofspatialorganizationincavesitesRouzaud (1997) proposed applicable criteria for the analysis of human use of caves.His approach is basedonthe criteria of illumination andorientation. As such,Rouzauds work is incorporated into the concept of spatial organization presented inthis article. The aforementioned condition of the human eye in the darknessprovidesatangiblebasisforthecalculationof different light zonesandchambertypes. Different chambers canbe connectedtoone another via a pathnetworkcomprising differentlines of communication andconnection points.In somecaves,e.g., Fontanet, gures are situated within clambering passages inside the cave(Delteiletal.1972),whichprovidesevidencethatthemodeofmovementandtheplacement of guresare probably linked.Therefore, the modeof movement shouldbeincludedas animportant featureof thisanalysis. Bymeasuringtheavailablespace at locations used by humans, the maximum number of people that could havestayedforthesametimeinthesameplaceshouldbeestimated.In summary, a tool kit for the analysis of spatial organization incave sitescomprisesvetools:lightzone, chambertype, pathnetwork, modeofmovement,andavailablespace.Trimmel (1965)hasreferredtothelight-darkboundaryasthenal point tobereachedbysunlight. Rouzaud(1997)referstothissamepointasthebeginningofthedarkzone. FollowingRouzaud, thereisafurtherspherebetweendaylightandthedarkzonetheso-calledhalf-shadezone(Fig.1).Theboundariesbetweenthedifferent zones can be ascertained, when possible, on the basis of illuminancemeasurements. Anilluminance of under 0.01lxnolonger permits anyformofJArchaeolRes1 3orientationormovement. Thus, it wouldappearreasonabletoplacethetransitionfromhalf-shadezonetodarkzoneat this particular point. Open-air illuminanceuctuates between2,000lx(overcast sky) and100,000lx(clear sky) (Griefahn1996). Since these values are directly inuenced by the position of the sun as well asby climatic factors, the maximumilluminance during adverse conditions, i.e.,overcast sky, is takenas theboundarybetweendaylight andtheaforementionedhalf-shade zone; this is around 2,000lx. Due to many variables, however, thegeneral transitional zonebetweenthesetwounitsmust retainacertainexibility.The aforementioned zones can serve only as approximations of potential conditions.Theclassicationof chamber typesisbasedonthelevel of illuminationbyasingle tallow lamp (Fig.2). Approximately 130 tallow lamps were found in the caveof Lascaux(Delluc andDelluc 1979), andone can imagine other scenarios ofillumination. Webasedour calculationonminimumlightingbyasingleperson.Gray-brown walls of caves do not expose the optimumreectivity. In thatenvironment, the illuminance of 0.25lxof a tallowlampcanlight anaveragedistance of 4m. We dene this distance as the radius withinwhichcontrolledmovement wouldhavebeenpossibleunder Paleolithicconditions byonesingletallow lamp. Narrow chambers stand out in particular because they are smaller than4mwide, so that both walls are adequately illuminated at the same time.Accordingly, wedenemedium-sizedchambersasmeasuringbetween4and8mwide; in these chambers at least one wall is reached by light at any one time. In widechambers just one wall can be illuminated. In low chambers the ceiling can serve asareector and, irrelevant of chamber width, canprovidequitesufcient light. Itfollows that there are only two chamber types for which illumination and orientationcan be problematic: inmedium-high and wide-high chambers. These chambersdaylight zonehalf-shade zonedark zoneFig.1 Lightzones(afterRouzaud1997):Daylightzone,half-shadezone,anddarkzoneJArchaeolRes1 3compel the visitor to be near one of the two walls in order to use its surface to reectlight.Thedifferent chamberswithinacaveareconnectedtooneanother viaapathnetwork that is made up of different lines of communication and connection points.Withrespectto the lines of communication, we differentiatebetweensidepassagesandpassageways.Whereasa sidepassagehasanexplorativecharacter,at thesametimeit isalsotheonlywayout. Withapassageway, thereisalwaysat least onealternativewayback.Differentlinesofcommunicationintersectandareconnectedviaconnectionpoints, crossings, junctions, deadends, andentrances. Acrossingconnects at least four different lines of communication, a junction at least three. Forus,adeadendisaspecialtypeofconnectionpoint;suchapathstopswithanendpieceinsteadofbeingconnectedtothepathnetwork.Decisive for the exploitation of caves is the mode of movement. Trimmel (1965)alsoconsideredthisaspect.Accordingly,differencescanbeascertainedinthewaypathsandconstrictedpathscanbenavigated. Whereaspathscanbecomfortablynegotiated, constricted paths can be passed only in a crawling position. Thisdifferentiationistakenupandincorporatedintothesystematicsofouranalysisofspatial organization. In addition to walking and crawling, climbing is anotheralternativemodeofmovementthatweconsider.Thespaceavailableinacaveisofparticularsignicanceforitsanthropogenicusage.AccordingtothearchitectNeufert(1951),ahumanrequiresasurfacespaceof 0.77m2when walking, and 1.75m2when lying down. If we add to this a value toreectacertaindegreeoffreemovingspace, thenwearriveatagureofaround2m2of spacethat isrequiredby eachhuman(e.g.,Pager1989). Withthisvalue inmind, wecalculatethemaximumnumberofpeoplethat couldhavestayedinthesameplaceat thesametime. Westress, however, that this valueshouldnot beunderstood as a precise unit of measure, as cave size can never be measured< 4m< 4mnarrow-low medium-low wide-lownarrow-high medium-high wide-high> 4m4 - 8m > 8mFig.2 ChambertypesbasedonthelevelofilluminationbyasingletallowlampJArchaeolRes1 3accurately owing to individual chamber structures. Therefore, we propose theapplication of a scale by which the available space is divided in steps of ve people(10m2=5people,20m2=10people,30m2=15people, andsoon).Caveartinitscontext:ArstapplicationofaspatialorganizationanalysisinthecavesitesofBedeilhac, Fontanet,andLePortel(Arie`ge)Inthis rst applicationof spatial organizationanalysis, caves were chosenthatfullled certain criteria. First, they featured not only cave art but also published dataonconcentrationsof other prehistoricremains. Inaddition, theconcentrationsofnds from each cave were culturally and chronologically coherent. Furthermore, thecavesareingeographical proximitytooneanothersothat theirgeographical andcultural uniformity is ensured. Three cave sites in Arie`ge in the southwest ofFranceBedeilhac, Fontanet, andLe Portelmet these conditions (Fig.3). Wehaveusedonlypublished dataandareawareofthedecitsandproblemsthatarisefrom such an approach. One major problem is the sloppy documentation of the sitesinpublications. Manydiscoveriesanddetailsfromeachsiteremainunpublished.Thereisagreat needfor apreciseanddetailedstandardof publicationfor suchimportant sites. Otherwise, a serious scientic usage of published data by thescienticcommunityisimpossible.BedeilhacInvestigationsat Bedeilhac(Fig.4) werebeguninthemid-19thcentury, rst byGarrigoufrom1861to1864, followedbyRobert from1941to1953(Malvesin-Fabreetal.1953),andsince1990 bySauvet(2007)andBarbaza,fromwhomonlypreliminaryreportsarecurrentlyavailable(Barbazaetal. 1996).ThecaveartfromBedeilhacis described asexceptionallyhomogeneous and,onthe basis of its stylistic traits, is attributed to the middle Magdalenian (Vialou 1986).AccordingtoVialou, thereare76guresat Bedeilhac; most areincolor(73.7%)andare foundthroughout the entire cave. Engravings are featuredinjust threeconcentrations: in the Salle terminale, the Galerie des modelages, and in the GalerieVidal.Thefrequentuseofblackandredsuggeststhatthegureswereintendedtobeseen; thisdoes not applytothe18engravings. Amongthe76guresare25animal representations, 7are indeterminate elements, and44are abstract signs.Animal representations are dominatedbybison(n=18) but alsoinclude horse(n=4), ibex (n=2), and deer (n=1). In addition to some solitary gures,Bedeilhacboastsfour concentrationswithdrawingactivitiesintheGalerieVidal(n=37), intheGaleriedesmodelages(n=16), intheSalleterminale(n=12),andtheDiverticuleauxbisons(n=3).TheGalerieprincipaleandtheDiverticuleaux bisons are the only localities in which gures solely of bison and abstract signsare found. The Galerie principale is signicant not just for this reason; it is the onlygalleryinwhichcaveartappearsscattered, i.e., thereisnoconcentrationxedononeparticularpoint.JArchaeolRes1 3Atleast11concentrationsofarchaeologicalobjectsandtwosingleoccurrencesare known in Bedeilhac (Barbaza et al. 1996; Malvesin-Fabre et al. 1953).Unfortunately, the level of information known about each concentration varies, andtheycanbecomparedonlywithrestrictions. Theinformationrangesfromsimple,supercial, jotteddownnotestomodernexcavations. Concentrations1-11followthe numbering by Malvesin-Fabre et al. (1953). The list was supplemented with theconcentration of archaeological objects in the Galerie Vidal (12), which wasexcavated by Barbaza in the middle 1990s (Barbaza et al. 1996). The following canbe derivedfromthe artifact inventories that have beenpublishedfor Bedeilhac(Barbazaetal.1996;Malvesin-Fabreetal.1953):veconcentrations(1,4,10,11,and12)haveevidenceofsubstantialactivitieswithblankproduction,formaltools,and the remains of hunting preparation. Five other concentrations (3, 5, 6, 7, and 9)have primarily formal tools without corresponding reduction sequences; theParisToulouse12345678910Fig.3 Franco-Cantabrian Europe showing the cave sites mentioned: Bedeilhac (1), Chauvet (2), Cussac(3), Fontanet (4),La Garma(5), Lascaux(6), Pergouset (7),LePortel (8),Les Trois-Fre`res(9), andTucdAudoubert(10)JArchaeolRes1 3activities carried out are unknown qualied activities. Concentration 2 lackssufcient documentation. Projectiles are found in 10 of the 11 concentrations(Barbazaet al. 1996; Malvesin-Fabreet al. 1953). Eventheverysmall inventorywith unknown qualied activities is characterized by projectiles. Except forconcentration 12, in addition to projectiles, mobiliary art is a dening characteristicoftheconcentrationswithsubstantialactivities.Inthesmallerconcentrationswithunknownqualiedactivities, caveartisfoundirregularly.Due tothe extremelylarge dimensions of the entrance area andthe straightcourse of the rst 150m, the boundary between the daylight and half-shadezone isnot distinct. The half-shade zone extends approximately to the junction to theGalerieVidal. AllotherchambersandgalleriesinBedeilhacareinthedarkzone.The illumination available for orientation in the cave is uncomplicated. In theGalerieprincipale, one must be nearone of the two walls because the ceiling is toohightoreectthelightofthetallowlamps.ThisalsoholdstruefortheGaleriededroite, thesecondentrancetothecave. Theother twogalleries anddiverticula,whichbranchofftotheside,aresonarrowandlowthatorientationissimple.Itispossible to stand and walkcomfortablyinallpartsof theBedeilhaccave.The onlyexceptionsaretheDiverticuleauxbisonsandtheGaleriedesmodelages, bothofwhicharepassagesinwhichonehastocrawl. Theonlyhindranceisaclimbingpassagewayinthe rst part of the Galerie Vidal that formedwhenthe ceilingcollapsedandtheremainingrockdebrisblockstheaccess.The path network in Bedeilhac consists mainly of side passages that terminate asdeadendsandjunctions.OnlytheGaleriedesmodelagesoffersthepossibilityofapassageway. WhenenteringtheGalerieprincipale, itispossibletoorientyourselfalongone wall andonthe wayout youcanorient yourself alonganother wall?substantialdarkhalf-shadedaylightconsumptionunknown qualifiedundocumenteddrawingnarrow-highmedium-highwide-highnarrow-lowmedium-lowwide-lowhuman activity light zone chamber type>100 side passagepassagewayjunctioncrossingdead-endwalkingcrawlingclimbingpeople (n)path network mode of movement available space1234567/9101112Galerie principaleSalle terminale>100?>100 ?>100>100805>1005Galerie principaleGalerie des modelagesDiverticule aux bisonsGalerie VidalGalerie de droiteentranceentrance50mBdeilhac>100drawing locationsupply locationdrawing locationdrawing location with substantial activityundocumentedsupply locationdrawing location with substantial activityFig.4 Distribution of human activities with structured notes about their spatial organization in Bedeilhac(numbers refer to concentrations)JArchaeolRes1 3(Simonnet 1976). These dead ends all branch off the central side passage. There arenocrossingsinBedeilhac.People began using the cave in the middle Magdalenian, based on stylistic,typological, andabsolute-chronologicalarguments(Vialou1986). Atthattime,allareasof thecavewereexploredandused. Inadditiontotheconcentrationswithsubstantial activities andthe others withunknownqualiedactivities, there areanotherveconcentrationswithdrawingactivities[GalerieVidal, Diverticuledesmodelages, Salle terminale (two concentrations), and Diverticule aux bisons](Sauvet 2007), for a total of 16 concentrations. Adenite synchrony of theseconcentrations cannot be assured due to the lack of chronological resolution.However, their cultural-chronological unity is very clear (Vialou 1986). Most of theconcentrations(74.3%) areintheunlit areaof Bedeilhac, whichunderscorestheinterest of stayinginthedeeper partsof thecave. Darknesswasobviouslynot ahindrance. Exceptionsaretwoconcentrationswithsubstantial activities(1and4)and two with unknown qualied activities (3 and 5); these are still in the half-shadezone.Theconcentrationsaredifferentlydistributedinthepathnetworkof thecave.Whereas concentrations withsubstantial activities are at junctions (1and4), indeadends(10and11), andinasidepassage(12), thosewithunknownqualiedactivitiesoccur onlyat junctions. Incontrast, concentrationswithdrawingactiv-ities occur in dead ends (Diverticule aux bisons and Salle terminale), in sidepassages(GalerieVidal), andinpassageways(Diverticuledesmodelages); junc-tions were avoided for drawingactivities. The placement of the concentrationswithsubstantial activities was orientedtowarddifferent factors: sufcient light,goodaccess, andsupervisionoftheentrance(1and4); yet proximitytocaveartalsowas desired(Galerie Vidal [12] andSalleterminale [10/11]). Furthermore,the concentrations with unknown qualied activities in the cave systemareorientedtowardthecentral sidepassage, theGalerieprincipale, yet alsotowardthejunctionstogallerieswithdrawingactivities. PerhapsthefunctionofsomeoftheseconcentrationsinBedeilhaccanbebetterdenedduetotheirlocationinthepathnetworkandtheprojectilesfoundthere. Concentrations6,7,and9arefoundat the junctionof the central pathwayaxis goingtoa concentration withonlydrawingactivities. Insomeoftheirinventoriesthereareseveral projectiles. Theirfunctionas a huntingweaponcannot initiallyexplaintheir presence ina cave.Perhapstheybelongtothebasictechnical equipment, requireextramaintenance,haveahighvalue, or simplyanother unknownfunction. Incomparisonwiththeother tools, the projectiles are foundfrequentlyinthe listedconcentrations andperhapshavespecialimportance.Theconcentrationswithsubstantialactivitiesarefoundinlocationsthat haveroomformorethan100people; thisisalsotrueforthe concentration that has both substantial and drawing activities in the Salleterminale. Alocation with comparable activities in the Galerie Vidal (12) hasroomfor80people,asdoestheconcentration(6)withunknownqualiedactivity;others(7and9)needlessspace. Theconcentrationswithonlydrawingactivities(Diverticule aux bisons and Diverticule des modelages) are in much smallerJArchaeolRes1 3chambers that have space for only ve people. The concentrations with onlysubstantial activities tend to be near the entrances, i.e., closer to daylight; thelocationswithdrawingactivitiesarefarther awayfromdaylight.FontanetInFebruary1972theGalerieWahlwasdiscoveredintheFontanetcave(Fig.5),afewkilometerssouth ofTarascon-sur-Arie`ge (Delteil et al. 1972).For the rst timesince the early discoveries at the beginning of the 20th century and the catastrophichandlingofthe Lascaux cave,whichwasdiscoveredin1940,anintact,completelypreserved concentration of prehistoric cave art, archaeological objects, and imprintswasfound;itwasanexceptionalnd.The cave art inFontanet is extremelyhomogeneous and, onthe basis of itsstylistic features, belongs to the middle Magdalenian (Ministe`re de la Culture 1984;Vialou1986). AccordingtoVialou(1986)thereare224gures, ofwhich77areengravingsand147arecolordrawings. Thefact that two-thirdsoftheguresarecoloredsupportstheassumptionthat theyweremadetobelookedat. Ofthe224gures, 142 areundened andmiscellaneous gures, 45 areabstract gures,31areanimals, and6areanthropomorphs. Amongtheanimal gures, bison(n=6)arethemost frequent. Therealsoarehorses(n=4), ibex(n=4), anddeer(n=2).Withintherst50moftheGalerieornee,thecaveartformsaclearconcentrationGrande salleGalerie profondeGalerie orne1 225< 252525entranceGalerie IGalerie IIGalerie IIIGalerie IVGalerie VABCCamarin5555?entranceblocked entranceLe Portel50m50mFontanetGalerie Wahldrawing location with consumption activitydrawing location with consumption activitydrawing locationdrawing locationdrawing locationdrawing location with consumption activitysupply locationsupply locationFig.5 Distribution of human activities with structured notes about their spatial organization in FontanetandLePortel(numbersandlettersrefertoconcentrations)JArchaeolRes1 3ofdrawingactivities.Elsewhere,inadditiontoisolated,sporadicgures,thereisaconcentrationofcaveartintheGalerieprofonde.Twoconcentrations of archaeological objects fromFontanet were mentioned(Vialou 1986), and another was listed by Clottes (Ministe`re de la Culture 1984), but nodetails were published. The following, however, can be summarized from the listedpublications: There are traces of re and a diverse faunal spectrum (Vialou 1986) butonlyafewlithicartifacts. Blankswerenot produced. Thus, nosubstantial activ-ities tookplaceinFontanet; rather, supplies wereconsumedthere(consumptionactivities). In the back area of the cave, two lithic artifacts were found on the surface ofthe oor of the cave (Delteil et al. 1972), indicating that there were not only the twoconcentrations in Fontanet but other isolated occurrences of lithic artifacts.According to the information from Vialou (1986), daylight reaches up to the rstsmallbendinthe300-m-longgallery. Thebendislocated40mfromtheoriginalentrance. It canbe assumedthat light conditions didnot change abruptlyfromthehalf-shadezonetothedarkzone. Therest of thecaveis inthedarkzone.Notwithstanding the actual range of the half-shade zone, the illumination fororientation in Reseau Wahl is uncomplicated.Only when cutting across the Grandesalledoesonehavetostayclosetooneofthewalls. Otherwise, thepathwidthof3-5m requires only a tallow lamp to light at least one of the cave walls. In all partsof the Reseau Wahl, it is possible to walk upright and comfortably. The onlyexceptionsaretheclimbingpassageandthecrawlingpassagebetweentheGrandesalle and the back part of the gallery. The path network here consists of only a singlesidepassagethat terminatesinadeadend. Sincethereisnopassagewayandnojunction, there are noalternatives for a different wayout. Inspite of the highstandarddeviationsof theabsolutedating, theprehistoricremainsfromFontanetforma cultural unit andare classiedas middle Magdalenian(Ministe`re de laCulture 1984; Vialou 1986). This corresponds to the results of the spatialorganization analysis of the cave. There is a clear center of prehistoric activities thatislocatedwithintherst50mofthecave.Over90%ofthecaveartiswithinthisarea, whichispartiallyinthehalf-shadezoneandpartiallyintheborderingdarkzone. Due to the simple structure of the Galerie Wahl in Fontanet, there is only onecentral sidepassage, which terminatesin a dead end. The climbing passage and thesingle crawlingpassage were not especiallymarked. The deadendwas not ofspecial interest. However, isolated gures, traces on the cave oor, and thediscoveryoftwoisolatedlithicartifactsareevidencethat peoplewerealsointhebackpart of thecave. Bothconcentrations withconsumptionactivities formtheclosureof theconcentrationwithdrawingactivities. Thereis roomfor approxi-mately 25 people in front of the different gures. Proximity to the cave entrance wasimportant for all activities. In sum, Fontanet has a simple pattern. The spacebetweendaylightandtheGrandesallewasusedasanactivitycenter. TheGrandesallewasavoided, andthebackpartofthecavewasonlyreconnoitered.LePortelAmid the Plantaurel limestone massif is the fossil cave system of Le Portel (Fig.5).Approximately30mbelowthis cavesectionis thesystemthat is activetoday.JArchaeolRes1 3Le Portel has hardly been investigated since its discovery in 1908; there are only theexcavations byNoulet, Vezian(Vezian1955), andrecordingof thecaveart byBeltranetal. (1967)andDauvois.Vialou (1986) counted 138 gures in Le Portel, including 64 animals, 41 abstractand28miscellaneous gures, and5anthropomorphs. Amongtheanimals, horse(n=26) and bison (n=23) are dominant. Deer (n=4), ibex (n=1), sh(n=1), and 8 undened animals also are represented. The stylistic features as wellas the continuous completion of the monochromatic drawings in eitherred or blackarguefor their homogeneity. Theyareclassiedas middleMagdalenian(Vialou1986). The cave art is fairly uniformly distributed in all parts of the cave; there is nodistinctcenter.Differencesbetweentheindividualgalleriescanbeseenamongthedominantmotifs. InGalerieIII(GalerieRegnault), horseisdominant. Incontrast,bisonisdominantinGalerieIV(GalerieBreuil)(Ministe`redelaCulture1984).At the beginning of the 20th century, four concentrations of archaeologicalobjects of the middle Magdalenian were found in Le Portel during prospection work(Vezian 1955). The rst concentration at the entrance to the Galerie de droitereceivednofurtherattentionintheliterature; duetolackofdata, it couldnot beincluded here. The available material on the artifact nds in the other threeconcentrationsissimilar andisclassiedasmiddleMagdalenian(Vezian1955).The presence of blanks, formal tools, and the remains of hunting preparation in twooftheartifact inventoriesindicatesubstantial activities(concentrationsAandC).The formal tools and faunal remains in concentration Bindicate consumptionactivities.Today, allofthegalleriesandchambersareinthedarkzone. Articiallightisneededfor orientationandmovement everywhere inthe cave, yet it is easytoilluminate the entire cave. Except for parts of Galerie V (Grande salle), the paths aresonarroworlowthatilluminationposesnoproblem.Incontrast,movementinLePortel is hindered by various small climbing passages. These are found repeatedly inGalerie III and Galerie IV. Upright movement is possible everywhere else.Especially notable is the Camarin in Galerie IV, a small chamber that can be enteredonlybycrawling. Inthepathnetwork, theGalerieI(GalerieJeannel), GalerieV,andGalerieII (GalerieJammes) formthecentral axis, andbothGalerieIII andGalerie IV junction off from there. There is no crossing in Le Portel. All of the sidepassagesterminateindeadends.Noneofthechamberscanserveasapassageway.There are different opinions on the uniformity and distribution of cave art withinthecave. Whereasthecaveart issubclassiedbyBreuil andLeroi-Gourhanintoearlytomiddle Upper Paleolithic andMagdalenianensembles (Ministe`re de laCulture 1984), Vialou(1986) stresses theuniformityof theprehistoric remains.According to the published artifact spectra of the different concentrations ofarchaeological objects (Vezian 1955), concentrations Aand Chave traces ofsubstantial activities; there are no drawing activities.The topographic locationnearthe possible original entrance ensures that concentration Chad natural light,controlledaccess, andquickentrancetothecavesystem, allbasicnecessitiesthatare ensured through positioning in Le Portel. The function of concentration A is noteasilydeduced. Entryintothecaveviathesteepandnarrownortheastentranceisdifcult. Althoughthereisalackofnaturallight, concentrationAmayhavebeenJArchaeolRes1 3placed so that the basic necessities could be achieved in a comfortable place.According totheartifactspectrum,theactivitiescarriedoutinconcentrationBcanbeclassiedasconsumptionactivities. Drawingactivitiesalsowerecarriedoutatthe same location. In the path network of Le Portel, this spot is located directly at ajunctioninthesecondquarter of thecavesystem. Except for individual spotsinGalerie II and Galerie V, the available space in Le Portel is limited basically to vepeople. Thelistedexceptionsclearlyhavemorespace(c.50people). Throughthedistribution of the concentrations with exclusively substantial activities, it ispossibletocontrol accesstotheentirecavesystem. Thisalsoholdstrueif bothentrances to the cave were open at the same time. The cave art is relativelyuniformlydistributedinthethreegalleries,eachofwhichterminatesinadeadendand is as far away from the entrance as possible. Thereis no distinct center.Due tolimited available space, the cave art is not intended to be viewed by larger groups atthesametime. Thus, theintensiveuseof color does not correspondtoalargeraudience.ConclusionsandperspectivesOur concept for the analysis of spatial organization of cave sites follows anintegrativeapproach.Allitemsfoundinwell-preservedcavesarehandledwiththesame care. Cave art, archaeological objects, and imprints are seen as equally rankedtraces of human activities and are classied using standard archaeological methods.Thespatialpatterningofhumanactivitiesinthecaveisanalyzedbyadenedtoolkit for spaceanalysisas usedtodayinurbanplanning. Thisstructuredapproachopens a neweldof investigation; a consistent nomenclature anddenitionoffeatures as light zone, chamber type, path network, mode of movement, andavailablespacecomplement intuitiveandindividualisticdescriptions. Therecon-structedhumanactivitiesarethenevaluatedintheirtopographicalandarchaeolog-ical contexts. The proposed set of features is easy to handle, even when drawn onlyfrompublished data. Adirect comparative analysis of cave sites is possible.Althoughthis broadapproachbears theriskof ignoringindividual, ne-graineddetailsofeachcave,ithastheadvantageofilluminatingpatternsinthehumanuseofcaves.In the three caves investigated here, 29 individual concentrations wererecognizedanddividedintotwobasiccategories: substantial activity(n=7)andqualied activity (n=21). One concentration could not be classied. The qualiedactivities comprise 13 drawing, 3 consumption, and 5 unknown. The relativepositions of theseconcentrations inthecaves combinetoproduce22locations.Projecting them onto maps of the caves produces four broad groups of locations thatare characterized rst by kind of human activity and second by spatial feature (lightzone, chamber type, path network, mode of movement, and available space)(Fig.6).(1)Supplylocations,wheresubstantialactivitiesarecarriedout,areinthehalf-shadeanddarkzones.Theselocationscanbereachedeasilybecausetheyarenearthe entrance. Access to the entire cave systemcan be controlled fromtheseJArchaeolRes1 3locations.(2) Absolutedarknessisimportantfordrawinglocationswithsubstantialactivities.Althoughsubstantialactivitiesarerecorded, theyhavelittleeffectonthepositioningwithinthecave. (3)Drawinglocationswithconsumptionactivitiesarevariablypositionedinthecaves.Theselocations,whichshowthelowestproleofdenition,arerecordedinthe half-shadeanddarkzones. (4)Drawinglocationsarethoseareaswherethereisnoevidenceofanyactivitiesotherthandrawingandnoarchaeological objects are found. They are positioned in the dark zone, provide littlespace, andarefoundinpassagewaysorsidepassagesthatterminateindeadends.Thatfourcave-specicgroupsoflocationsweregeneratedfromthe22differentlocationstestiestosomekindofpatterninginthehumanuseofcavesandtheirspatial arrangement inthecavelandscape. What isparticularlyremarkableistheorientationofthelocationsalongtherespectivepathnetworkofthecaves(Figs. 4and 5). Complete control (physical and symbolic) over all movements in the cave isachievedthroughthesupplylocations. Locationsthat exclusivelydisplaydrawingactivities are restricted to the dark zone. Very narrow spaces that accommodate onlyafewpeopleat atime were selectedfor this. Locations inwhichdrawingandsubstantial or consumption activities were carried out have too few data at present toallow the recognition of dened patterns of their positioning in the path network. Ofparticular importance is the observation that substantial activities were carried out inBedeilhacandLePortel. Thesesubstantial activitiesaremacroscopicidentical tothoseat open-airsites. Theyindicatethe provisioning of basic needs. Due to a lackof data, it is not possible to carry out calculations on the length of stay or group size.However, thistypeoflocationinthecavemakessenseonlyifpeopleintendedtostayinsidethecaveforalongerperiodoftime. Itwouldbehelpfultoknowhowlongthesupplyoffoodwouldhavelastedandtowhat extent theproductionandimprovement of tools, whichwere used outside later,were carried out inthe caves.Thesedatacouldindicatehowoutsideactivitiescarriedout afterthestayinthe25 525 55 >100>100>10080 >100Bdeilhac: -Fontanet: 1/2Le Portel: BBdeilhac: Div. aux bisons/ Gal. des modelagesFontanet: Gal. profonde/ Gal. orneLe Portel: Gal. II/ Gal. III/ Gal. IVBdeilhac: 1/4Fontanet: -Le Portel: A/CBdeilhac: 3/5/6/7/9Fontanet: -Le Portel: -Bdeilhac: 2Fontanet: -Le Portel: -Bdeilhac: 10/11/12Fontanet: -Le Portel: -drawing locationsupply locationundocumenteddrawing location withsubstantial activitiesdrawing location withconsumption activities??human activityn43351237light zone chambertypepath network mode of movementavailable spacecave site: concentration location type6 6 5 4 51 1 1 1 3 1 23 2 4 4 212 23 2 2 3 21 1 12 2 2 3 21 1 1 12 4 4 4 4 21 1 1 1 1?substantialdarkhalf-shadedaylightconsumptionunknown qualifiedundocumenteddrawingnarrow-highmedium-highwide-highnarrow-lowmedium-lowwide-lowhuman activity light zone chamber type>100 side passagepassagewayjunctioncrossingdead-endwalkingcrawlingclimbingpeople (n)path network mode of movement available spaceFig.6 Combinationof humanactivities andthe structurednotes about their spatial organizationinBedeilhac,Fontanet, andLePortelJArchaeolRes1 3cavewere included in the overall planning process. This additional information isnecessary for a holistic interpretation of cave use and would provide indications fortheinterpretationofthecaveart.TherearenosupplylocationsatFontanet.Thesupportoftheinternalactivitiesfrom a location within the cave was not necessary. Due to the small number of cavesin our analysis, we do not have a counterpart for this pattern. A reasonable argumentforthespatialsettingatFontanetwouldbetheshortdistancefromthelocationstotheexterior. Averypreliminarylookat asimilar situationinLascauxseems tosupportthisassumption.Up to now, only the gures themselves were used in the interpretation of cave art.Our integrativeconcept includes all prehistoricremains for adetailedanalysis ofhumanbehaviorincaveswithrockart. Thecavesunderstudyindicateabalancedrelationship between drawing activities and substantial or consumption activities. Suchabalancedrelationshipmight indicatethat thecavewas surveyedcompletelybyPaleolithic users todevelopa mental mapof the space andanthropomorphize itafterwards. The creation of various recurrent locations throws light on the extraordinarycapabilities of Paleolithic users in spatial recognition of the caves. The patterning givesevidence that a kind of master plan for residence in caves was in operation.Our results should be regarded as a rst basic step to provide a spatial frameworkfor the interpretation of cave art locations in their context. It would be a tremendousadvanceifresearchersofcaveart wouldagreetousethesamenomenclatureandwould describe their sites in precise reports that allowfurther scientic applications. ItisobviousthatdifferentconceptsormasterplanswerefollowedbythePaleolithicusers. Aprolonged stayina cave interior supportedby supplylocations as inBedeilhacandLePortel affords strategies other thanthevisits inFontanet. Themeaning of this varying behavior is still open to debate. We have to admit that in thepresent stateofresearch, eventheritual characterofthisbehaviorisquestionable(AriasCabal2009,p.287).Formoredetailedresearchweneedtoincludefeaturesother than space, as there are, e.g., the tangible and the intangible, visibility and views,sound, and public or private access (Bahn 2003). The art panels themselves are packedwith detailed quantitative information that is waiting to be unpacked by researchers.Spatial statistical analysis of cave art provides access to this information (Classen andZimmermann 2003). The combination of spatial organization of cave architecture asproposedherewithspatialpatterningandorganizationofartpanelslinkedbyGIStools will provide an excellent database in the near future. This approach should allowaninterpretationofPaleolithicrockartthatabandonsnarrativeightsoffancyandtouches down on empirical terrain.Acknowledgments TheGermanScienceFoundation (DFG) supportedthisresearchproject from2001to2004.P.G.Bahn, B.Hayden,T.Lenssen-Erz,K.Kindermann,H. Riemer,theanonymous reviewers,andtheeditorsgaveusefuladvice.K. SchneiderandL.ClaretranslatedthetextintoEnglish.ReferencescitedAriasCabal, P. (1999). LaGarma(Kantabrien/ Spanien): EiszeitlicheWandkunst undWohnplatzeineiner verschlossenen Hohle. Jahrbuch des Romisch-Germanischen Zentralmuseums Mainz 46: 320.JArchaeolRes1 3Arias Cabal, P. (2009). Rites inthedark?Anevaluationof thecurrent evidencefor ritual areas atMagdaleniancavesites.WorldArchaeology41:262294.Arias Cabal, P., Gonzalez Sainz, C. G., and Moure Romanillo, A. (2003). Unterirdischer Raum,Wandkunst undpalaolithische Strukturen: Einige Beispiele der Hohle La Garma (Spanien). InPastoors,A.,andWeniger,G. C.(eds.),HohlenkunstundRaum:Archaologischeundarchitekton-ischePerspektiven, JanvanderMost,Dusseldorf, pp.2946.Aujoulat, N. (2002). La grotte ornee de Cussac - le Buisson-de-Cadouin (Dordogne): premie`resobservations.BulletindelaSociete PrehistoriqueFrancaise99:129137.Bahn, P. G. (2003). Location, location:WhatcanthepositioningofcaveandrockartrevealaboutIceAge motivations? In Pastoors, A., and Weniger, G. C. (eds.), Hohlenkunst und Raum:ArchaologischeundarchitektonischePerspektiven, JanvanderMost,Dusseldorf, pp.1120.Barbaza, M., Lacombe, S., andGalop, D. (1996). Bedeilhac-et-Aynac: GrottedeBedeilhac, GalerieVidal. In Ministe`re de la Culture (ed.), Bilan Scientique de la Region Midi-Pyrenees 1995,Ministe`redelaCulture,Toulouse,pp.3132.Bartenbach, C., and Witting, W. (2009). Handbuch fur Lichtgestaltung: Lichttechnische undwahrnehmungspsychiologischeGrundlagen, Springer,Vienna.Beaune, S. A. de (1987). Lampes et godets au Paleolithique, Centre National de la RechercheScientique,Paris.Begouen, H. (1926). Lart mobilier de la caverne du Tuc dAudoubert (Arie`ge). Jahrbuch furPrahistorischeundEthnographischeKunst(IPEK)2:219228.Begouen, R., andClottes, J. (1981). ApportsmobiliersdanslescavernesduVolp(Enle`ne, LesTrois-Fre`res, Le Tuc dAudoubert). In Ministerio de Cultura (ed.), Altamira Symposium: Madrid-Asturias-Santander,15-21octobre1979, MinisteriodeCultura,Santander,pp.157188.Begouen, R., Fritz, C., Tosello, G., Clottes, J., Faist, F., andPastoors, A. (2007). Les Magdaleniensmodelaientaussilargile.LesDossiersdelArcheologie324:3037.Begouen, R., Fritz, C., Tosello, G., Clottes, J., Pastoors, A., and Faist, F. (2009). Le sanctuaire secret desbisons:Ilya14000ans, lartet laviedesMagdaleniensdanslacaverneduTucdAudoubert,Somogy,Paris.Beltran, A., Robert, R., andGailli, R. (1967). LaCuevaBedeilhac, Facultadde Filosof a yLetras,Zaragoza.Binford, L. R. (1983). In Pursuit of the Past: Decoding the Archaeological Record, Thames and Hudson,NewYork.Bonsall,C., andTolan-Smith,C. (eds.)(1997).TheHumanUseofCaves, Archaeopress,Oxford.Classen, E., and Zimmermann, A. (2003). Raumliche Statistik, soziale Netzwerkanalyse undRaumverstandnis.InPastoors,A.,andWeniger,G. C. (eds.),HohlenkunstundRaum:Archaolog-ischeundarchitektonischePerspektiven, JanvanderMost,Dusseldorf,pp. 91104.Clottes,J.(ed.)(2001).LaGrotteChauvet:lartdesorigines, Seuil,Paris.Delluc, B., andDelluc, G. (1979). LEclairage. InLeroi-Gourhan, A., andAllain, J. (eds.), Lascauxinconnu, CentreNationaldelaRechercheScientique,Paris,pp.121142.Delteil, J., Durbas, P., and Wahl, L. (1972). Presentation de la galerie ornee de Fontanet (Ornolac-Ussat-les-Bains,Arie`ge).BulletindelaSociete PrehistoriquedelArie`ge27:1120.Eggert,M.K. (2001).PrahistorischeArchaologie:KonzepteundMethoden, A. Francke,Tubingen.Fritz, C., and Tosello, G. (2007). Le secteur de la Salle Hillaire et de la Salle du Crane: Diversite, styles etdatationdelart paleolithiquedanslagrotteChauvet. InFloss, H., andRouquerol, N. (eds.), Leschemins de lart aurignacien en Europe. Actes du colloque 2005 dAurignac, Editions Musee-ForumAurignac,Aurignac, pp.393408.Geneste, J. M. (ed.) (2005). Recherches pluridisciplinaires dans la grotte Chauvet, SocietePrehistoriqueFrancaise,Paris.Griefahn,B. (1996).Arbeitsmedizin, Thieme,Stuttgart.Harten,H. U. (1997).PhysikfurMediziner:EineEinfuhrung, Springer,Berlin.LeGuillou,Y. (2005). Circulationshumainesetoccupationdelespacesouterraina` lagrotteChauvet-Pont-dArc.InGeneste,J.M.(ed.), Recherches pluridisciplinaires dans la grotte Chauvet,SocietePrehistoriqueFrancaise,Paris,pp.117134.Lenssen-Erz, T. (2001).Gemeinschaft-Gleichheit-Mobilitat. FelsbilderimBrandberg, Namibia,undihreBedeutung:GrundlageneinertextuellenFelsbildarchaologie, Heinrich-Barth-Institut,Koln.Leroi-Gourhan,A. (1965).Prehistoiredelartoccidental, Mazenod,Paris.Leroi-Gourhan,A. (1971).Prehistoiredelartoccidental, 2nded.,Mazenod,Paris.Lorblanchet, M.(1995).Lesgrottesorneesdelaprehistoire:nouveauxregards, Errance,Paris.JArchaeolRes1 3Lorblanchet, M. (2001). La grotte ornee de Pergouset (Saint-Gery, Lot: un sanctuaire secretpaleolithique, EditionsdelaMaisondesSciencesdelHomme,Paris.Malvesin-Fabre,G.,Nougier,L.R.,andRobert,R.(1953).L0occupationmagdaleniennedelagrottedeBedeilhac (Arie`ge) et decouverte d0un nouveau gisement dans la galerie Vidal. Bulletin de la SocietePrehistoriquedelArie`ge8:2048.Ministe`re de la Culture (ed.) (1984). Lart des caverns: atlas des grottes ornees paleolithiques francaises,ImprimerieNational, Paris.Neufert, E. (1951). Bauentwurfslehre. Grundlagen, Normen und Vorschriften uber Anlage, Bau,Gestaltung, Raumbedarf undRaumbeziehungen: Maefur Gebaude, Raume, EinrichtungenundGeratemitdemMenschenalsMaundZiel, Tempelhof,Berlin.Pager,H.(1989).BeobachtungenzurAuswahlderFelsbildstellenimBrandberg.InPager,H.(ed.), TheRock Paintings of the Upper Brandberg: Part I - Amis Gorge, Heinrich-Barth-Institut, Koln,pp.4980.Pasda, C. (2004). Hotel Grnland: Human Use of Caves and Rock Shelters in West Greenland,Archaeopress,Oxford.Pastoors, A., and Weniger, G. C. (eds.) (2003). Hohlenkunst und Raum: Archaologische undarchitektonischePerspektiven, JanvanderMost,Dusseldorf.Rouzaud, F. (1977). Elements pour la paleospeleologie des Pyrenees centrales au paleolithiquesuperieur,Memoirepourlobtention dudiplome,LEcoledesHautesEtudesenSciences Sociales,Toulouse.Rouzaud, F. (1996). Lapaleospeleologie: unemethodedetudedesgrottesprehistoriqueset paleonto-logiques. In Delporte, H., and Clottes, J. (eds.), Pyrenees prehistoriques: arts et societes, ComitedesTravauxHistoriquesetScientiques,Paris,pp.143148.Rouzaud, F. (1997). La paleospeleologie ou: lapproche globale des documents anthropiques etpaleontologiquesconservesdanslekarstprofond.Quaternaire8:257265.Sauvet, G. (2007). LagrotteorneemagdaleniennedeBedeilhac. LesDossiersdelArcheologie324:5461.Sieveking, A. (1997). Cave as context in palaeolithic art. In Bonsall, C., and Tolan-Smith, C.(eds.), TheHumanUseofCaves,Archaeopress,Oxford,pp. 2534.Simonnet, R. (1976). La grotte de Bedeilhac. InClottes, J. (ed.), UISPP, IXCongre`s: Livret-guideExcursionA5,Universite deNice,Nice,pp.8387.Streich, B. (2005). StadtplanunginderWissensgesellschaft:EinHandbuch, VerlagfurSozialwissens-chaften,Wiesbaden.Trimmel, H. (ed.)(1965). FachworterbuchfurKarst-undHohlenkunde(SpelaologischesFachworter-buch), LandesvereinfurHohlenkundeinWienundNiederosterreich,Wien.Vezian, J. (1955). Les foyers magdaleniens de la grotte du Portel (Arie`ge). Bulletinde laSocietePrehistoriquedelArie`ge910:1332.Vialou, D. (1986). Lart des grottes en Arie`ge magdalenienne, Centre National de la RechercheScientique,Paris.Villeneuve, S., andHayden, B. (2007). Nouvelle approche de lanalyse ducontexte des gurationsparietales. In Beaune, S. A. de (ed.), Chasseurs-cueilleurs: comment vivaient nos ancetres duPaleolithiquesuperieur,CentreNationaldelaRechercheScientique,Paris,pp.151160.Wrede, V. (1996). Fremde Welten unter unseren Fuen: Was sind Hohlen. In Rosendahl, W., and Krause,E. B. (eds.), ImReich der Dunkelheit. Hohlen und Hohlenforschung in Deutschland, EditionArchaea,Gelsenkirchen,pp.1014.BibliographyofrecentliteratureAriasCabal, P., andOntanonPeredo, R. (eds.) (2004). Lamateriadel lenguajeprehistorico: el artemueblepaleolticodeCantabriaensucontexto, GracasCalima,Santander.AriasCabal,P.,OntanonPeredo,R.,AlvarezFernandez,E.,CuetoRapado,M.,Garcia-Monco Pineiro,C., and Teira Mayolini, L. C. (2008). Falange grabada de la galer a inferior de la Garma: aportacionalestudiodelartemobiliardelMagdalensiensemedio.VELEIA2425:97129.Aubry, T., Mangado, J., Sampaio, J. D., and Sellami, F. (2002). Open-air rock art, territories and modes ofexploitationduringtheUpperPaleolithicintheCoaValley(Portugal).Antiquity76:6276.Aujoulat,N. (2004).Lascaux:lageste,lespaceetletemps, Seuil,Paris.JArchaeolRes1 3Bahn,P.G.,Franklin,N.,andStrecker,M.(eds.)(2008).RockArtStudies:NewsoftheWorld,OxbowBooks,Oxford.Barbaza, M. (1994). Bedeilhac-et-Aynac: grottedeBedeilhac. InMiniste`redelaCulture(ed.), Bilanscientiquedelaregionmidi-Pyrenees1993, Ministe`redelaCulture,Toulouse, p.28.Barbaza, M. (1997). Bedeilhac-et-Aynac: grottedeBedeilhac. InMiniste`redelaCulture(ed.), Bilanscientiquedelaregionmidi-Pyrenees1996, Ministe`redelaCulture,Toulouse, pp.2829.Barbaza,M. (1997). Premie`resdatations14CdesniveauxarcheologiquesdelagalerieVidal, grottedeBedeilhac(Arie`ge).BulletindelaSociete PrehistoriquedelArie`ge52:3344.Barbaza, M., andLacombe, S. (1995). Bedeilhac-et-Aynac: grotte de Bedeilhac. InMiniste`re de laCulture (ed.), Bilan scientique de la region midi-Pyrenees 1994, Ministe`re de la Culture, Toulouse,pp.2829.Barbot, L. (2007). Le cerveau createur dart: approche neuroscientique de la creation artistique. In Floss,H., and Rouquerol, N. (eds.), Les chemins de lart aurignacien en Europe: Actes du Colloque 2005dAurignac,EditionsMusee-ForumAurignac,Aurignac,pp.259274.Bicho, N. F., Carvalho, A. F., GonzalezSainz, C. G., Sanchidrian, J. L., VillaverdeBonilla, V., andStraus, L. G. (2007). The Upper Paleolithic rock art of Iberia. Journal of Archaeological Method andTheory14:81151.Bosinski, G. (2003). Lacaverneparticipante: HohlenraumundHohlenbilderimJungpalaolithikum. InPastoors,A.,andWeniger,G. C.(eds.),HohlenkunstundRaum:Archaologischeundarchitekton-ischePerspektiven, JanvanderMost,Dusseldorf, pp.2128.CantalejoDuarte, P. (2006). LaCuevade Ardales: arte prehistoricoy ocupacionenel PaleolticoSuperiorestudios, 1985-2005, ServiciodePublicacionesCentrodeEdicionesdelaDiputaciondeMalaga,Malaga.Chippindale, C., andNash, G. (2004) (eds.). PicturesinPlace: TheFiguredLandscapeof Rock-Art,CambridgeUniversityPress,Cambridge.Chippindale, C., andTacon, P. S. (eds.) (1998). TheArchaeologyof Rock-Art, CambridgeUniversityPress,Cambridge.Classen, E., Kindermann, K., and Pastoors, A. (2009). Djara: Cave art in Egypts Western Desert. Archeo-Nil19:4766.Clottes,J.(2008).CaveArt, PhaidonPress,NewYork.Conkey, M. W. (1992). Lessitesdagregationet larepartitiondelart mobilier, ou: Ya-t-il dessitesdagregation magdaleniens? In Rigaud, J. P., Laville, H., and Vandermeersch, B. (eds.), LepeuplementMagdalenien:paleogeographiephysiqueethumaine, Comite desTravauxHistoriquesetScientiques,Paris,pp.1925.Conkey, M. W. (1997). Beyondart andbetweenthecaves: Thinkingabout context intheinterpretiveprocess. InConkey, M. W., andSoffer, O. (eds.), BeyondArt: PleistoceneImageandSymbol,UniversityofCaliforniaPress,SanFrancisco,pp.343367.CorchonRodr guez,M.S.(1997).Lacornichecantabriqueentre15000et13000ansBP:laperspectivedonneeparlartmobilier.LAnthropologie101:114143.Cremade`s, M. (1996). Lart mobilier pyreneen: analogies technologiques et relations intersites. InDelporte, H., andClottes, J. (eds.),Pyreneesprehistoriques:artsetsocietes, Comite desTravauxHistoriquesetScientiques,Paris, pp.367380.Delluc,B., andDelluc, G.(1984).Lascaux:artetarcheologie, EmmanuelLeymarie,Perigueux.Desbrosse, R., andThevenin, A. (eds.) (2007). Artset culturesdelaprehistoire: hommagesa`HenriDelporte, Comite desTravauxHistoriquesetScientiques,Paris.Domingo Sanz, I., Fiore, D., and May, S. K. (eds.) (2008). Archaeologies of Art: Time, Place and Identity,LeftCoastPress,WalnutCreek,CA.Errico,F.de,Vanhaeren,M.,SanchezGoni,M.F.,Grousset, F.,Valladas,H.,andRigaud,J.P.(2001).Les possibles relations entrelart des caverneset lavariabilite climatiquerapidedeladernie`reperiode glaciaire. In Barrandon, J. N., Guibert, P., and Michel, V. (eds.), Datation: XXIe rencontresinternationalesdarcheologieetdhistoiredAntibes, EditionAPDCA,Antibes,pp.333347.ForteaPerez, F.J.(ed.)(1993).Laproteccionyconservaciondelarterupestrepaleolitico, Astur-Graf,Oviedo.Fritz, C., Tosello, G., andSauvet, G. (2007). Groupes ethniques, territoires, echanges: lanotiondefrontie`redanslart magdalenien. InCazals, N., GonzalezUrquijo, J. E., andTerradas, X. (eds.),Frontie`res naturelles et frontie`res culturelles dans les Pyrenees prehistoriques, Ediciones delaUniversidaddeCantabria, Santander,pp. 165182.JArchaeolRes1 3Gonzalez, R. (2001). Art et espace dans les grottes paleolithiques cantabriques, Jerome Million,Grenoble.Hodgson,D.(2006).Understandingtheoriginsofpaleoart:Theneurovisualresonancetheoryandbrainfunctioning.Paleoanthropology2006:5467.Hodgson, D.,andHelvenston, P.A.(2006). Theemergenceof therepresentationofanimals inpaleoart:Insightsfromevolutionandthecognitive,limbicandvisualsystemsofthehumanbrain. RockArtResearch23:340.Lenssen-Erz, T. (2002). Art for arts sake intherockart research: Onthe ethics of archaeologicaldocumentation. InLenssen-Erz, T., Tegtmeier, U., andKropelin, S. (eds.), Tidesof theDesert -Gezeitender Wuste: Contributions totheArchaeologyandEnvironmental Historyof AfricainHonourofRudolphKuper, Heinrich-Barth-Institut, Koln,pp. 549558.Lenssen-Erz, T., andNeubig, J. (2003). AugenblickundEwigkeit, RaumundDiskurs. Artefaktederprahistorischen Kunst Namibias und die Arteplage in Murten, Expo.02 in der Schweiz. In Pastoors,A., and Weniger, G. C. (eds.), Hohlenkunst und Raum: Archaologische und architektonischePerspektiven, JanvanderMost,Dusseldorf, pp.8190.Leroi-Gourhan, A., andAllain, J. (eds.) (1979). Lascauxinconnu, Centre National de la RechercheScientique,Paris.Lewis-Williams, J. D. (1998). Wrestling with analogy: A methodological dilemma in upper Paleolithic artresearch.InWhitley,D.S.(ed.), Reader in Archaeological Theory. Post-processual and CognitiveApproaches, Routledge,London,pp.157178.Lewis-Williams, J. D. (2002). The Mind in the Cave: Consciousness and the Origins of Art, Thames andHudson,London.Lorblanchet, M., and Bahn, P. G. (eds.) (1993). Rock Art Studies: The Post-stylistic Era or Where Do WeGofromHere? Papers PresentedinSymposiumAof the 2ndAURACongress, Cairns 1992,Oxbow,Oxford.Mellars, P. (2006). TheecologicalbasisofUpperPaleolithiccaveart. InMa lloFernandez, J. M., andBaquedano,E.(eds.),MiscelaneaenhomenajeaVictoriaCabrera:VolumenII,GracasAlgoran,Alcala deHenares,pp. 211.Montes Barqu n, R., MunozFernandez, E.,and MorloteExposito, J.M. (2004). Cueva Urdiales (CastroUrdiales,Cantabria):estudiogeo-arqueologicoyarterupestrepaleoltico, Quinzanos,Santander.Morton,J.(1999). Arrernte. InLee,R.B.,andDaly,R.(eds.), The Cambridge Encyclopedia of HuntersandGatherers, CambridgeUniversityPress,NewYork,pp.329334.Moure Romanillo, A. (1990). Relations entre art rupestre et art mobilier en region cantabrique. In Clottes,J. (ed.), Lart des objets au paleolithique. Tome 1: Lart mobilier et soncontexte, Laballery,Clamecy,pp.207218.Moure Romanillo, A. (1996). Art parietal, art mobilier et habitat: etat de recherche. In Thiault, M. T., andRoy, J. B. (eds.), Lart prehistorique des Pyrenees, Reunion des Musees Nationaux, Paris,pp.7279.Octobon, E. (1938). Art et magie dans la grotte de Bedeilhac (Arie`ge). Revue Anthropologique 48: 4154.Pastoors, A., andWeniger, G. C. (2004a). LacollectionWendel: archiveenimages delart parietalpaleolithique. The Wendel collection: A picture archive of ice age cave art. International NewsletterofRockArt40:2427.Pastoors,A.,andWeniger,G.C.(2004b).BilderimDunkeln:HohlenkunstderEiszeit.DieSammlungWendel,CD-ROM,NeanderthalMuseum, Mettmann.Pettitt, P. B., andBahn, P. G. (2003). Current problemsindatingpalaeolithiccaveart: CandamoandChauvet.Antiquity77:134141.Pettitt, P. B., Bahn, P. G., and Ripoll Lopez, S. (eds.) (2007). Palaeolithic Cave Art at Creswell Crags inEuropeanContext, OxfordUniversityPress,Oxford.Pettitt, P. B., Bahn, P. G., and Zuchner, C. (2009). The Chauvet conundrum: Are claims for thebirthplace of art premature? In Bahn, P. G. (ed.), An EnquiringMind: Studies inHonor ofAlexanderMarshack, OxbowBooks,Oxford,pp.239262.Pigeaud, R. (2005). Unart detraces?Spontaneiteset premeditationssur lesparoisdesgrottesorneespaleolithiques. InVialou, D., Renault-Miskovsky, J., andPatou-Mathis, M. (eds.), Comportementdes hommes duPaleolithique Moyenet Superieur enEurope: territoires et milieux, DerouauxOrdina,Lie`ge,pp.177192.Pincon,G.,Fuentes,O.,andBourdier,C. (2007).Lasculptureenabri:unartpartage?LesDossiersdelArcheologie324:8691.JArchaeolRes1 3Ross, M. (2001). Emergingtrends inrock-art research: Hunter-gatherer culture, landandlandscape.Antiquity75:543548.Sacchi, D. (ed.) (2002). Lart paleolithiquea`lairlibre:lepaysagemodierparlimage, GraepandGeopre,Carcassone.Sauvet,G. (1990). Lessignesdanslartmobilier. InClottes, J. (ed.),Lartdesobjetsaupaleolithique,tome2:Lesvoiesdelarecherche, Laballery,Clamecy,pp.83100.Sauvet, G. (1992). Bedeilhac-et-Aynat: grotte de Bedeilhac. InMiniste`re de la Culture (ed.), Bilanscientiquedelaregionmidi-Pyrenees1991, Ministe`redelaCulture,Toulouse, pp.3133.Sauvet, G. (1993). Bedeilhac-et-Aynac: grotte de Bedeilhac. InMiniste`re de la Culture (ed.), Bilanscientiquedelaregionmidi-Pyrenees1992, Ministe`redelaCulture,Toulouse, pp.2526.Sauvet, G. (1994). Bedeilhac-et-Aynac: grotte de Bedeilhac. InMiniste`re de la Culture (ed.), Bilanscientiquedelaregionmidi-Pyrenees1993, Ministe`redelaCulture,Toulouse, pp.2931.Sauvet, G. (1995). Bedeilhac-et-Aynac: grotte de Bedeilhac. InMiniste`re de la Culture (ed.), Bilanscientiquedelaregionmidi-Pyrenees1994, Ministe`redelaCulture,Toulouse, pp.2931.Sauvet, G. (1996). Bedeilhac-et-Aynac:grottedeBedeilhac,salleterminale. InMiniste`redelaCulture(ed.), Bilan scientique de la region midi-Pyrenees 1995, Ministe`re de la Culture, Toulouse,pp.3334.Sauvet, G. (1997). Bedeilhac-et-Aynac: grotte de Bedeilhac. InMiniste`re de la Culture (ed.), Bilanscientiquedelaregionmidi-Pyrenees1996, Ministe`redelaCulture,Toulouse, pp.2931.Sauvet, G. (2001). Bedeilhac-et-Aynat: grotte de Bedeilhac. InMiniste`re de la Culture (ed.), Bilanscientiquedelaregionmidi-Pyrenees2000, Ministe`redelaCulture,Toulouse, pp.1920.Sauvet, G. (2005). Bedeilhac-et-Aynat: grotte prehistorique de Bedeilhac. In Ministe`re de la Culture (ed.),Bilanscientiquedelaregionmidi-Pyrenees2002, Ministe`redelaCulture,Toulouse,pp. 23.Sauvet, G., Fritz, C., and Tosello, G. (2007). Lart aurignacien: emergence, developpement,diversication. In Cazals, N., Gonzalez Urquijo, J. E., and Terradas, X. (eds.), Frontie`res naturelleset frontie`res culturelles dans les Pyrenees prehistoriques, Ediciones de la Universidad de Cantabria,Santander,pp.319338.Swartz,B.K.,andHurlbutt,T.S.(1994).Space,placeandterritoryinrockartinterpretation.RockArtResearch11:1322.Vialou,D. (1998).Lartdesgrottes, EditionsScala, Paris.Villeneuve, S. N. (2008). Looking at caves from the bottom-up: A visual and contextual analysis of fourPaleolithic painted caves in southwest France (Dordogne). MA thesis, Department of Anthropology,UniversityofVictoria,Victoria.Whitley,D. S.(ed.)(2001).HandbookofRockArtResearch, AltaMiraPress,WalnutCreek,CA.Zilhao,J.(2007).Theemergenceofornamentsandart:Anarchaeologicalperspectiveontheoriginsofbehavioralmodernity.JournalofArchaeologicalResearch15:154.JArchaeolRes1 3