A Comparative Analysis of Châtelperronian and Protoaurignacian Blade Core Technology Using Data Derived from 3D Models

Conference Paper (PDF Available) · March 2017with 243 Reads
Conference: Computer Applications and Quantitative Methods in Archaeology (CAA) International Conference, At Atlanta, Georgia
Abstract
This study uses data extracted from 3D models to compare the Châtelperronian and Protoaurignacian stone tool industries, which are at the center of the debate surrounding the nature and extent of interactions between Neanderthals and anatomically modern humans (AMHs) approximately 40,000 years ago. Our data are derived from 3D scans of blade cores associated with one Protoaurignacian and two Châtelperronian assemblages from two archaeological sites (Roc-de-Combe and Les Cottés). We will use these data to test hypotheses about technological similarities and differences between these two entities by making statistical comparisons of artifact attributes, which are either difficult or impossible to quantify using traditional methods. We will examine the utility of different approaches including the digital measurement of edge angles, and the characterization of artifact shape using elliptical fourier analysis. We will interpret our results within the framework of previous qualitative observations made about these technologies, as well as the larger research question of if and how Neanderthals and AMHs may have exchanged technological knowledge. This presentation will also touch on larger issues including the use of close-range photogrammetry for morphological analyses of artifacts, the pros and cons of using 3D-derived data compared to traditional methods, and the quality of data required to conduct a study of this kind.
Figures - uploaded by Samantha Thi Porter
Author content
All content in this area was uploaded by Samantha Thi Porter
No caption available
… 
No caption available
… 
No caption available
… 
No caption available
… 
A Comparative Analysis of
Châtelperronian and Protoaurignacian
Blade Core Technology Using
Data Derived from 3D Models
Samantha Porter
Morgan Roussel
Marie Soressi
CAA International Conference
March 14-16, 2017
Atlanta, GA
1
Châtelperronian
~45-40 kya cal BP
Protoaurignacian
~42-40 kya cal BP
Les Cottés
Roc de Combe
Archaeological Context
2
Research Questions / Design
Can we use 3D scanning to quantify observed qualitative
differences between Châtelperronian and Protoaurignacian
lithic technology?
oWhat are the implications for the relationship between Neanderthals and
anatomically modern Homo sapiens?
Three quantitative measures investigated
oAngle between core surfaces
oAngle between core axes
oPlatform shape (Elliptical Fourier Analysis)
3
Scanning Methods
4
3D Models
Châtelperronian Protoaurignacian
CTS RDC
5 cm
5
Angle Between Core Surfaces
Lycett and von Cramon-Taubadel, 2015
Y5-1574
Châtelperronian
Y6-1584
Protoaurignacian
6
Angle Between Core Surfaces
7
Results - Angle Between Core Surfaces
35
45
55
65
75
85
95
35
45
55
65
75
85
95
RDC_CHÂT CTS_CHÂT CTS_PROTO
8
Results - Angle Between Core Surfaces
35
45
55
65
75
85
95
RDC+CTS_CHÂT CTS_PROTO
There is a statistically
significant difference in the
angle between core
surfaces between the
combined Châtelperronian
and the Protoaurignacian
sample.
Performed a two-tailed
t-test assuming equal
variance:
p=0.0102
9
Core / Platform Shape
Châtelperronian: Hypothesized Angular Central Tendency
Roussel, Soressi, and Hublin, 2016
top view
10
Core / Platform Shape
Protoaurignacian: Hypothesized Rounded Central Tendency
After Bon, 2006
top view
11
Core / Platform Shape
12
Core / Platform Shape
13
Core / Platform Shape
14
Core / Platform Shape
SHAPE
Iwata and Ukai, 2002 15
Results - Platform Shape
‐0.5
‐0.4
‐0.3
‐0.2
‐0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
‐1.2 ‐1 ‐0.8 ‐0.6 ‐0.4 ‐0.2 0 0.2 0.4 0.6
PC1
PC2
60.57% of
variance
31.06% of
variance
16
Results - Platform Shape
‐0.25
‐0.2
‐0.15
‐0.1
‐0.05
0
0.05
0.1
0.15
0.2
‐0.5 ‐0.4 ‐0.3 ‐0.2 ‐0.1 0 0.1 0.2 0.3 0.4 0.5 0.6
PC2
PC3
31.06% of
variance
3.54% of
variance
Further investigation needed…
17
Angle Between Core Axes
Châtelperronian Protoaurignacian
Hypothesized Central Tendencies
top view
18
Angle Between Core Axes
19
Angle Between Core Axes
20
0
5
10
15
20
25
30
35
40
45
RDC_CHÂT CTS_CHÂT CTS_PROTO
Results - Angle Between Core Axes
0
5
10
15
20
25
30
35
40
45
21
Results - Angle Between Core Axes
There is a statistically
significant difference in
core axes angles for the
combined Châtelperronian
and Protoaurignacian
sample.
BUT…
Performed a two-tailed
t-test assuming equal
variance:
p=0.0498
0
5
10
15
20
25
30
35
40
45
RDC+CTS_CHÂT CTS_PROTO
22
Results - Angle Between Core Axes
There is NOT a statistically
significant difference in
core axes angles between
the Les Cottés
Châtelperronian and Les
Cottés Protoaurignacian
sample.
Performed a two-tailed
t-test assuming equal
variance:
p=0.1145
0
5
10
15
20
25
30
35
40
45
CTS_CHÂT CTS_PROTO
23
What Does Quantitative
Analysis Tell Us?
Lithic artifacts are not biological organisms, and are highly
variable
What is driving our results?
oIs this ‘real’ evidence of similarities in knapping behavior?
oAre we measuring / quantifying things ‘the wrong way’?
We need to take into consideration factors outside of technology
oRaw material scarcity
oTypes of raw materials available
oProficiency of the knapper
oAllometry
oObserver error
24
Future Directions
Increase the sample size
Sample differently
oSeparate core types
oSeparate blade vs bladelet cores
oExclude more outliers
oInclude size in shape analysis
Look at different measures
oFurther investigate EFA data (e.g. run discriminant analysis)
oUse a different definition of cross section
oVector analysis and / or indices of convexity (Bretzke and Conard 2012)
25
Conclusions
Overall, our quantitative analyses supported previous qualitative
observations, but…
Metrics must be interpreted within the context of their
connectedness to human behaviors
3D scanning opens up new research avenues
oWe can make new types of systematic observations on artifacts
oWe can make certain observations more accurately
oWe can collaborate more easily
oWe can use 3D scans for many studies without having to travel to collect
additional raw data
26
Jean-Jacques Cleyet-Merle, André Morala, and the rest of the
staff of the Musée National de Préhistoire
Jacques Pelegrin
Glen Favretto and Steffen Schatz
Dana Winkler
University of Minnesota Doctoral Dissertation Fellowship
Please direct inquiries to Samantha Porter – port0228@umn.edu
Thank You
27
Bon, F. (2006). A brief overview of Aurignacian cultures in the context of Middle-to-Upper transitional
industries. Towards a Definition of the Aurignacian. IPA, Lisbon, pp. 133e144. Trabalhos de Arqueologia, (45).
Bretzke, K., & Conard, N. J. (2012). Evaluating morphological variability in lithic assemblages using 3D models of
stone artifacts. Journal of Archaeological Science, 39(12), 3741-3749.
Dibble, H. L., & Bernard, M. C. (1980). A comparative study of basic edge angle measurement techniques. American
Antiquity, 857-865.
Iwata, H. and Y. Ukai (2002) SHAPE: A computer program package for quantitative evaluation of biological shapes
based on elliptic Fourier descriptors. Journal of Heredity 93: 384385
Lycett, S. J., & von Cramon-Taubadel, N. (2015). Toward a “quantitative genetic” approach to lithic variation. Journal
of Archaeological Method and Theory, 22(2), 646-675.
Porter, S. T., Roussel, M., & Soressi, M. (2016). A simple photogrammetry rig for the reliable creation of 3D artifact
models in the field lithic examples from the early upper Paleolithic sequence of Les Cottés (France). Advances in
Archaeological Practice, 4(1), 71-86.
Roussel, M., Soressi, M., & Hublin, J. J. (2016). The Châtelperronian conundrum: Blade and bladelet lithic
technologies from Quinçay, France. Journal of human evolution, 95, 13-32.
References
28
Core / Platform Shape
Châtelperronian Protoaurignacian
CTS RDC
29
Core / Platform Shape
30
Core / Platform Shape
31
Core / Platform Shape
32
Core / Platform Shape
33
Core / Platform Shape
34
Porter,Roussel,andSoressi(2017).AComparativeAnalysisofChâtelperronianandProtoaurignacian
BladeCoreTechnologyUsingDataDerivedfrom3DModels.CAAInternationalConference,
Atlanta,Georgia.March14‐16,2017.
Slide1
Helloeveryone.Thanksforcomingonthislastafternoonoftheconference.
Slide2
Startingwithsomecontext,todayI’llbetalkingabouttwostonetoolindustries:theChâtelperronian
andtheProtoaurignacian.Whatisinterestingaboutthesetwoindustriesisthattheyexistedin
approximatelythesametimeinapproximatelythesameplace,butareassociatedwithtwodifferent
hominingroups:theChâtelperronianwithNeanderthalsandtheProtoaurignacianwithanatomically
modernhumans.TherehasbeenahugeamountofdebateonwhetherthesetwogroupsmetinEurope
approximately40,00yearsagoandhowtheymayhaveinteracted.Agreatdealofthisdebatehas
surroundedtheirlithictechnology.BoththeChâtelperronianandProtoaurignacianareblade‐based
technologies.Insimpleterms,thismeansthepiecesofstonetheywereproducingtomakeintotools
tendedtobelongandskinny,ortwiceaslongastheyarewide.Manyhaveinterpretedthisasbeingan
improvementoverprevious,socalledflakebasedproductionsystems,whichwereusedby
NeanderthalsinthisregionbeforetheChâtelperronian.Thisbegsthequestion,didNeanderthalscome
upwiththismore“advanced”technologythemselves,orweretheysomehowinfluencedbyincoming
Homosapiens?
I’mdefinitelynotgoingtoanswerthattoday,buthopefullythisprojectwillbeonestepingettingalittle
closertofiguringthingsout.ThereasonI’minthissessionisthatwechosetoinvestigatethese
questionsaboutlithictechnologyusing3DscansoflithiccoresfromtwositesinFrance:RocdeCombe,
whichcontainsChâtelperroniandeposits,andLesCottéswhichincludesbothChâtelperronianand
Protoaurignacianlayers.
Slide3
Thequestionbeingaskedinthisparticularstudyiscanweuse3Dscanningtoquantifyobserved
qualitativedifferencesbetweenChâtelperronianandProtoaurignacianlithictechnology?Lateron,we
hopetointerpretinthecontextoftheselargerquestionsaboutgroupinteraction,culturaltransmission,
andsoon.
WetheninvestigatedsimilaritiesanddifferencesbetweentheChâtelperronianandProtoaurignacian
usingthreemetrics,whichI’lldescribeinamoment.
Slide4
Togenerate3Dmodelsforthisproject,weusedAgisoftPhotoScan.Imageswerecapturedwiththehelp
ofthisrig,whichisdescribedinourpaperinAdvancesinArchaeologicalPractice.Imagesweretaken
overthecourseofthreeyears,between2014and2016.Overtimethissetupdidchangeprogressively
forthebetter.Iamwellawarethedatacollectionproceduresusedforthisprojectarenotideal,
neverthelesstheywerequitecontrolled.Forexample,inthispaperpublishedinAntiquitywecompared
thecontrolphotogrammetrysetupIdevelopwithamoreexpedientrigwiththecameraonauto,
withoutatripod.Aswetalkedaboutthismorninginthissession,byusingsomesimplecontrolsyoucan
getamuchhigherqualitymodel.
Slide5
Hereyoucanseethe613Dmodelsweusedinoursubsequentanalyses.Intotal,ourcollectionof3D
modelsincludes185cores.Weendedupexcludingmanyartifactsfromthisanalysis,includingcores
thatwereabandonedordamagedearlyonorwereofuncertainprovenience.
Slide6
ThefirstmeasureI’mgoingtodiscussistheanglebetweenwhatwecalltheplatformsurfaceandthe
flakingsurfaceofthecore.Thisanglehasabigeffectontheshapeofproductsextractedfromagiven
core,andneedstobetightlycontrolledbyknappers.Thismeansthatdifferencesinthisanglebetween
industrieswouldbeastrongindicatorofknappersmakingdifferentchoices.Thesearesomeextreme
examples,butbasedonpreviousobservationswehypothesizedthatthisanglewouldbehigherinthe
ChâtelperronianandlowerintheProtoaurignacian.
Anglesonstonetoolsarenotoriouslydifficulttomeasurephysically.Whenthisdonewithagoniometer,
whichistraditional,thevariationinmeasuredvaluescanbeashighas20degrees.
Slide7
Bymeasuringanglesdigitally,wecangreatlyincreasetheprecisionandverylikelyaccuracyofour
measurements.Forthisstudy,thiswasaccomplishedusingGeomagicDesignX.Thepolygons
representingthetop,platformsurfaceofthecorewereselected,andusedtocreateabestfitplane.
Next,weselectedthepolyfacescorrespondingtothelastsuccessfulremovalonthecore,andused
thosetocreateanotherbestfitplane.Wethenmeasuredtheanglebetweenthesetwoplanes.
Slide8
Herearetheresultsofoursurfaceanglemeasurements.Ontheright,wehavethisrepresentedasbox
andwhiskerplot.VisuallywecanseethattheChâtelperroniansampledoesindeedappeartohavea
highervalueforthisanglemeasurements.
Slide9
WecombinetheChâtelperroniansamplefromourtwositesandcompartedittoourProtoaurignacian
datausingasimplet‐test.Andwegotastatisticallysignificantp‐valueof0.01.Thissupportsprevious
workthatindicatesthismeasureisdifferentbetweenthesetwoindustries.
Slide10
Thenextthingwewantedtoinvestigatewastheshapeofthecoreplatformsasviewedincrosssection
fromabove.Nottoboreyouwithtoomuchlithictechnology,butviatraditionallithicanalysisithas
beenproposedtheChâtelperronianknapperstendedtoworkcoresusingtwodistinctsurfaces.One
surfacewasworked,thenanother,backandforth.Thishypotheticallyproducesacorewithasomewhat
triangularcrosssection.
Slide11
Incontrast,Protoaurignacianknapperstheoreticallyworkedthecoresusingasinglemorerounded
surface.Thiswouldresultinaroundedcrosssection.Ourideawastoextractactualcrosssectionsfrom
our3Dmodels,whichwouldbedifficulttodoaccuratelywithout3Dscanning,andthencomparethe
crosssectionsforChâtelperronianandProtoaurignaciancoresusingellipticalFourieranalysis.
Slide12
Toobtain2DcrosssectionsforellipticalFourieranalysis,weagainusedGeomagicDesign.Weusedthat
samebestfitplanefrombefore,thenmoveditdowntoensureitwouldcutthroughthecore.Wethen
extractedacrosssectionbasedontheintersectionofthisplanewiththe3Dmodel.Wealsomarked
pointscorrespondingtotheextremesoftheflakingsurface,andaddedavectorbetweenthem.
Slide13
Wethenviewedthisfromabove,orthogonaltothebestfitplaneandextractedimagesbothofthe
entirecoreandthecoreoutlinewiththisadditionalreferencegeometryandthenturnedtheminto
thesetypesofblackandwhiteimages.
Slide14
Wedecidedtoonlyincludethepartoftheplatformcrosssectionboundedbythevectorbetweenthe
endsoftheworkingareainordertohoneinontheareasofthecorethatwerethefocusofthe
knappers.
Forthisparticularanalysisweonlyusedtheimagesofthecorecrosssection,buttheseimagesofthe
entirecorewillcomebackintoplaylateron.
Slide15
WeusedthepackageSHAPEbyIwataandUkaitoprocessourcrosssectionsandtoconducttheelliptical
Fourierandprincipalcomponentsanalysis.Fortheanalysesdiscussedinthispaper,scalewasnottaken
intoconsideration,andweusedtenharmonics.
Slide16
I’dcalltheseresultspreliminary,andIdon’twanttooverinterpretthings,butlookingatthefirsttwo
principalcomponentsgraphed,weseeasignificantamountofoverlap.PC1accountsfor60.5%ofthe
variance.Intermsofinterpretation,PC1appearstodescribewhetherornotthecrosssectionisskewed
toeithertherightortheleft.Crosssectionssuchastheseoftencanbeinterpretedasresultingfrom
errorsmadebytheflitknapper.Inthesecases,thestonefracturesinawaythatmakesitdifficultforthe
knappertoshapethecoreappropriatelygoingforward.Theseerrorsarenotnecessarilymorecommon
inonemodeofstonetoolworkingversusanother.
PC2accountsforanadditional31%ofthevariance,whichrelatestotherelativedepthoftheplatform
surface.Wehavenotyethadtimetotestthis,butthislikelycorrelateswithdifferentcore“types”,
whosedescriptionsIwillspareyoufornow.Inordertonotoverlylimitoursamplesizewedidnot
distinguishbetweendifferent”types”ofcoreshere.Thisdoesmeanendedupwithalargeamountof
morphologicalvariation.
Slide17
HerearePCs2and3.Again,weseealotofoverlap.However,theredoesappeartopotentiallybea
differenceonPC3intermsofthisquestionsofroundedvsangularplatformshapes.PC3accountsfora
muchlowerthreeandahalfpercentofavariance.ItdoeslookslikemoreChâtelperroniancorestendto
havemoreangularplatformcross‐sections.Thisdoesfitouroriginalhypothesisthatwasbasedon
qualitativeobservations.
Nevertheless,thequestionremains,isthisdistinctionmeaningfulanddrivenbydifferenceinlithic
technology?Unfortunately,wewereunabletorunmoreanalysesbeforethisconference,sotheanswer
willhavetowait.Again,thisprincipalcomponentonlyaccountsfor3%ofthevariationinthesample.
Furtherinvestigationintothismeasureincludingsomemorestatisticaltestswillbenecessaryto
determinewhatisdrivingthisvariation,andiftheuseofellipticalFourierisusefulforourparticular
researchquestions.
Slide18
Thelastthingwelookedatwastheanglebetweenwhatwearecallingthesymmetryaxisandretreat
axisofanartifact.Forthesefigures,youcanimaginethatyouarelookingatthesecoresfromabove.
Thisareaoutlinedbythedottedlinecorrespondstothesurfaceofthecorebeingworked.Inblue,you
havethesymmetryaxis,whichcorrespondstoeitherthelongestaxisofthecore,orinsomecasesan
axisperpendiculartothelongestcoreaxis.Inred,youhavetheretreataxisofthecore.Todetermine
theretreataxis,youfirstdrawalineacrossthecorebetweenthegreatestextentofthesurfacesthat
havebeenworked.Soinotherwordsweseeremovalsstartinghereandcontinuinguntilthispoint.The
retreataxisisdefinedasbeingperpendiculartothisline.
BasedontheinterpretationsandChâtelperronianandProtoaurignacianlithictechnologydescribed
earlier,weexpectedtheanglebetweenthesetwoaxestobehigherintheChâtelperronianduetoa
obliqueprogressionofthereduction.Weexpectedtheanglebetweenthesetwoaxestobelowerinthe
Protoaurignacian,withreductionoccurringsymmetricallyaroundtheaxisofsymmetry.
Slide19
Forthisanalysis,weusedthesameblackandwhiteimagesextractedfortheellipticalFourieranalysis.
ThesewerethenprocessedusingmacrosinImageJ.Wecreatedaverticallinethroughthecenterofthe
platformareaforreference,andsuperimposedabest‐fitellipseovertheentirecore.
Slide20
Theimageswerethensuperimposedononeanother,andtheanglebetweentheclosestaxisofthe
ellipseandthisdividinglineweremeasuredinImageJtothenearestdegree.
Slide21
Herearetheresultsofouranalyses.Wehavesomepotentialoutliers,butoveralltherawdatamatches
ourpredictionthatthisanglewouldbehigherintheChâtelperronian.
Slide22
Onceagainweranat‐testonthecombinedsampleandloandbeholdweonceagaingotasignificantp‐
value,whichappearstosupportourinitialhypothesis.But…
Slide23
ifweonlylookatthesamplefromLesCottesandexcludetheRocdeCombematerial,wedonotgeta
statisticallysignificantdifference.Whatdoesthismean?I’dsayrightnowit’shardtosay.
Slide24
ThereareafewthingsIcansayhowevertoputtheseresultsintocontext.First,stonetoolsarehighly
variable.Duetothewayinwhichtheyareproduced,Iwouldalsoarguethatstonecoresaremore
variablethenotherartifacttypesthatcommonlyundergosimilaranalyses,suchasceramicsfor
example.
Thismeansweneedtothinklongandhardaboutwhataredrivingourresults.Arethesenumberswe’ve
generatedreallyrepresentativeofknappingbehaviors?
Weneedtotakeintoconsiderationotherfactorsincludingtheabundanceandqualityofrawmaterials
inagivenarea,tohowgoodtheindividualswhomadetheseartifactsweregoodatflintknapping.We
alsoneedtoinvestigatewhetherartifactshapechangesasartifactsgetlargerorsmaller,alsoknownas
allometry.Andofcourse,weneedtotakeintoconsiderationpossibleerrorsmadebyus,theobservers.
Slide25
Intermsoffuturedirections,thereareanumberofavenueswecouldpursue.Inthelongterm,itwould
benicetogetaccesstomorematerialtoincreasethesamplesize.Shortertermusingthescandatawe
alreadyhavewecouldplaywithoursampleabit–forexamplewecouldseparatecoretypes,separate
corebasedonsize,excludeoutliers,oraddsizeintoourshapeanalysis.
Wecouldalsolookatdifferentmetricsandstatistics.Wecouldrunthingslikediscriminantanalyses.We
couldmaybelookatcrosssectionadifferentway.Therearealsoafewmethodswehaveyettotry,
namelyvectoranalysisandcalculatingindicesofcorecurvature,bothofwhichweredescribedinthis
paperbyBretzkeandConard.
Slide26
Inconclusion,wewanttoemphasizeonceagainthatinarchaeologymetricsmustalwaysbeinterpreted
inthecontextoftheirrelationshiptohumanbehaviors.Thisstudydidproducesomeinterestingresults
thatmaysupportinterpretationsthatwerepreviouslybasedonmostlyqualitativeobservations.
However,moreworkneedstobedonetoaccountformoreofthepreviouslymentionedfactorsthat
alsoeffectstonetoolmorphology.
Finally,webelievethisstudydemonstratestheutilityofcollecting3Ddataforarchaeologicalresearch.
Inourcase,3Ddataallowedustomakenewtypesofsystematicobservationsonartifacts,takecertain
measurementswithmoreprecision,andcollaboratewithcolleaguesonthesamematerialdespitebeing
ondifferentcontinents.Furthermore,ifwewanttorunfurtheranalysesontheseartifacts,wecan
continuetoqueryour3Dmodelswithouthavingtotraveltotheartifacts’physicallocation.
Slide27
Thankyou
Slides28+
Theseslidesarehereincasesomeoneasksmetomorefullyexplainthedataextractionprocedure
  • Article
    Laminar knapping is a system of methods through which a recurrent series of elongated and morphometrically repetitive blanks are obtained. The appearance of this method is associated with cognitive changes that foster different paleoanthropological discussions. This article assesses the implications of the 497D lithic assemblage of Cova Gran (Northeast Pre-Pyrenees, Iberia), whose technical principles place it within the laminar knapping sphere in which the production of blades, bladelets, and flakes are interspersed. Refitting dynamic approach allows establishing physical connections, reflecting volumetric reduction integrate sequences of sequential blanks organized by the principle of technical predetermination. Techno-typological, contextual, and chronometric attributes enable the integration of 497D within an initial phase of the Upper Paleolithic. This is an important point; although laminar knapping is widely dispersed in space and time, it marks the emergence of anatomically modern Homo sapiens associated with the Aurignacian (sensu lato), a tradition marking the beginning of the Upper Paleolithic. This concept has been challenged by the proposal that the Châtelperronian laminar technocomplex was made by Homo neanderthalensis. These opposing positions fostered the Middle-Upper Paleolithic “transition” debate. Technical traits identified through refitting facilitate examination of similarities and differences between 497D and the Proto/Early Aurignacian and Châtelperronian. Observations arising from refitting in 497D influence characterization of the knapping method and its possible correlation with one of these Early Upper Paleolithic technocomplexes. These inferences broaden the goals usually associated with blade knapping, and encourage reconsideration of the classical definition of the laminar system.
This research doesn't cite any other publications.