-
Research Article
The Importance of Chert in Central Anatolia: Lessons fromthe
Neolithic Assemblage at Catalhoyuk, TurkeyAdam Joseph Nazaroff,1,*
Adnan Baysal,2 and Yahya Ciftci3
1Department of Anthropology, Stanford University, Stanford, CA,
USA2Archaeology Department, Faculty of Letters, Bulent Ecevit
University, Zonguldak, Turkey3Maden Tetkik ve Arama Enstitusu,
Universiteler Mahallesi Dumlupnar Bulvar, Cankaya, Ankara,
Turkey
Correspondence*Corresponding author;
E-mail: [email protected]
Received15 June 2012
Accepted28 February 2013
Scientific editing by Steve Kuhn
Published online in Wiley Online Library
(wileyonlinelibrary.com).
doi 10.1002/gea.21446
This paper is a presentation of three sources of
artifact-quality chert in CentralAnatolia. A previous dearth of
research focused on locating and characterizingsuch raw material
sources has incorrectly colored our view of prehistoric eco-nomic
practices. To remedy this situation, we have conducted a survey of
var-ious locales within Central Anatolia to test for the presence
of artifact-qualitymaterials. We make use of Individual Attribute
Analysis (IAA) and Energy Dis-persive X-ray Fluorescence (EDXRF)
analysis to discriminate among these ma-terials, and suggest their
use by the Neolithic occupants of Catalhoyuk throughsimilar
analyses of artifactual materials. We argue that the presence and
char-acterization of these new sources allows us to better
understand the intrica-cies of Neolithic practices by illustrating
the ways in which the consumption ofthese materials was variably
entwined with the exploitation of other resources,as well as
embedded within social relations outside of Central Anatolia. C
2013Wiley Periodicals, Inc.
INTRODUCTION
In the Middle East, the variable distribution of local
re-sources has been viewed as one component for creat-ing contexts
for the development of a diverse Neolithicworld (Gerard, 2002;
Thissen, 2002; Hole, 2003; Asouti,2006). Case studies in Central
Anatolia have shownhow local practices created multiple
trajectories for dif-ferent social developments during the
Neolithic (Ger-ard, 2002; Hodder, 2011). While research has focused
onthe use of woodland catchments (Asouti, 2005; Fairbairnet al.,
2005), pastoral environments (Pearson et al.,2007), ground stone
(Baysal, 1998; Turkmenoglu et al.,2005), and clay deposits
(Doherty, 2008), considerablyless attention has been given to
understanding the con-sumption of chert, a sedimentary rock
comprised of mi-crocrystalline or cryptocrystalline silica,
generally in ex-cess of 90% (Luedtke, 1992; Lowe, 1999). Although
acomparatively minor component of many lithic assem-blages, chert
is ubiquitously present at Neolithic sites inCentral Anatolia
(Bezic, 2007). Understanding the distri-bution and use of chert
offers a more thorough perspec-tive on the role of diverse local
practices in shaping theNeolithic.
At the Neolithic site of Catalhoyuk (74006000 B.C.)(Figure 1) in
Central Anatolia, James Mellaart statedthat the fine-grained chert
materials used in object pro-duction originated from locales in
southeastern Anatoliaor northern Syria (Mellaart, 1967:213,
1975:103). Thistheory was not unreasonable given the state of
knowl-edge at the time of Mellaarts writing. However, in re-cent
years our understanding of Anatolian geology hasexpanded, and
reports of chert materials are abundantin the geologic literature.
So much so that archaeolo-gists currently working at Catalhoyuk
have proposed thatartifact-quality sources may exist in the Western
TaurusMountains much closer to Catalhoyuk than Mellaart
hy-pothesized (Bezic, 2007; Doherty et al., 2007; see
alsoOstaptchouk, 2011). These new ideas primarily relatethe
presence of radiolarian chert-bearing ophiolite forma-tions common
in regions adjacent to the Taurus Moun-tains. However, a review of
the regional geologic liter-ature indicates several ophiolite
deposits and radiolariancherts throughout the whole of Anatolia
(Waldron, 1984;Valor & Tunay, 1996; Bozkurt et al., 1997;
Tekin, 2002;Parlak & Robertson, 2004; Okay, 2008), suggesting
addi-tional locations that may have been of use to Neolithic
340 Geoarchaeology: An International Journal 28 (2013) 340362
Copyright C 2013 Wiley Periodicals, Inc.
-
NAZAROFF ET AL. NEOLITHIC CHERT AT CATALHOYUK, TURKEY
Figure 1 Map of the Konya Basin, with local chert sources and
Neolithicsites indicated (modified after Ylmaz, 2010, Figure
1).
populations. Other possible deposits of chert may ex-ist in the
areas surrounding Ankara (Balkan-Atl, 1994),and concentrations of
chert have been reported near theGoksu Valley (Reynolds, 2007). In
order to move beyondconjecture, intensive survey of possible raw
material lo-cales and thorough characterization of both
artifactualand source material is necessary.
To date, no study has focused principally on the surveyand
collection of chert raw materials in Central Anatolia.Research
begun in 2011 by the Anatolian Archaeologi-cal Raw Material Survey
(AARMS) has sought to rem-edy this problem. The explicit goals of
AARMS relate tolocating, mapping, and characterizing a variety of
rawmaterial sources of possible social and economic impor-tance to
populations inhabiting Anatolia in prehistory.Further, it is a
primary objective of AARMS to providethis information in
conjunction with studies of artifactassemblages in order to posit
ancient interactions withphysical and cultural landscapes as
evidenced throughbehaviors of direct and indirect material
procurement.In this paper, we limit our discussion to the
introduc-tion of two newmaterial sources in Central
AnatoliatheSuhut and Akdere Chert sourcesand the confirmationof a
third chert source previously thought to exist alongthe western
fringe of the Konya Basin (Bezic, 2007).Furthermore, we discuss the
visual and geochemicalcharacterization of these sources, and
utilize such data todiscriminate among them as best as possible.
When us-ing the term source, we speak of a spatially discrete
de-posit of a material type discovered during survey, and
anaccompanying visual and geochemical signature derivedfrom
systematic sampling at the locale. When we proposethat an artifact
be assigned to a geologic source, we sug-
gest that the visual and geochemical characteristics of
theartifact match those of the source to a greater or lesser
de-gree of certainty. We have followed a framework offeredby
Shackley (2008:197198; see also Luedtke, 1992) forlocating,
sampling, characterizing, and presenting the re-sults of our
study.
We begin this paper with a discussion of the charac-teristics of
chert consumption during the Central Anato-lian Neolithic period
(10,0006000 B.C.) (Ozbasaran &Buitenhuis, 2002), with
particular emphasis on work pre-viously conducted at Catalhoyuk.
Next, we broadly re-view the geology of Central Anatolia, which
aided us inisolating regions for survey. Following this, we
presentthe results of our survey and a description of locatedchert
sources. Finally, we use Individual Attribute Analy-sis (IAA) and
Energy Dispersive X-ray Fluorescence Anal-ysis (EDXRF) to match
artifact materials from the earlyNeolithic levels at Catalhoyuk to
located sources.
CATALHOYUK IN CONTEXT:BACKGROUND TO CHERT CONSUMPTIONIN
NEOLITHIC ANATOLIA
The Neolithic occupation of Catalhoyuk commenced atroughly 7480
B.C., and spanned approximately 1400years (Cessford, 2001; Table
I). The location of such alarge site at a substantial distance from
Neolithic cen-ters of domestication in the Middle East marks its
im-portance in long-ranging social and economic networks(Cauvin,
2000). However, it is likewise important to rec-ognize how
Catalhoyuk is a product of developmentswhich occurred in a more
localized historical, social,
Table I Catalhoyuk East Mound chronology for South Area
excavations
(modified after Love, 2012).
Mellaart Excavation Hodder Excavation Levels,
Levels South Area
6000 B.C. 0-III TP 6 Levels
T
S
R
Q
P
VIA O
VIB N
VII M
VIII L
IX K
X J
XI I
XII H
7400 B.C. Pre-XII G1
Hodder Level G2
Geoarchaeology: An International Journal 28 (2013) 340362
Copyright C 2013 Wiley Periodicals, Inc. 341
-
NEOLITHIC CHERT AT CATALHOYUK, TURKEY NAZAROFF ET AL.
environmental, and economic context (Thissen, 2002;Asouti,
2006:117; Hodder, 2011). While much effort hasbeen expended in
understanding the procurement andconsumption of other materials at
Catalhoyuk, less con-sideration has been given as to how the
acquisition anduse of chert operated alongside these other
practices.
The Neolithic chert assemblage at Catalhoyuk con-sists of more
than 4000 samples from excavations of theNeolithic occupation in
the south and north areas ofthe East Mound (Conolly, 1999). Brief
commentaries onthese materials have been offered by Bezic (2007),
Carteret al. (2005), and Conolly (1999). Many of these indicatethat
the use of chert was intimately entwined in a rangeof other
practices, and differed substantially from theuse of other chipped
stone materials, particularly obsid-ian (Bezic, 2007). For example,
elaborate bone-handledknives recovered at Catalhoyuk are made of
fine-grainedchert materials, and find no correlation in the
obsidian as-semblages (Conolly, 1999; Hodder, 2006). Chert is
rarelycached at Catalhoyuk, while obsidian is frequently de-posited
in such contexts. Ratios of obsidian and chertknapping debris also
suggest that some obsidian indus-tries focused on household
production, while chert wasoften brought to site in the form of
preforms and blanks(Bezic, 2007:82). At the nearby sites of Can
Hasan andSuberde, chert was used for the exclusive manufacture
of sickle blades. Variable ethnographic and
ethnohistoricaccounts have suggested that the processes
underwritingthe variable uses of such materials often relate to the
dif-ferential use of material sources (e.g., Gould et al.,
1971;McBryde, 1978, 1984, 2000; Binford & OConnell, 1984;Gould
& Saggers, 1986).
Five sources of chert are commonly discussed in
thearchaeological literature of Anatolia (Figure 2). These in-clude
ophiolite-related radiolarian chert deposits in theWestern Taurus
Mountains, similar radiolarian chertssouth of Beysehir Golu and
surrounding the moderncity of Mersin, and high-quality chert
sources in east-ern Anatolia adjacent to Kahramanmaras and
Gaziantep(Sakcagoz) (Bezic, 2007; Doherty et al., 2007; see
alsoReynolds, 2007). Other than the latter two deposits, theexact
locations, depositional character, and macroscopic,petrographic,
and chemical characteristics of these mate-rials is unknown.
Bezic (2007:7076) has offered an excellent reviewof chert
consumption during the Central AnatolianNeolithic. In Central
Anatolia, chert consistently com-prises less than 5% of chipped
stone assemblages, whilesites in the Lake District and
Urfa-Diyarbakir Regionoften have lithic assemblages principally
composed ofchert during the Neolithic (up to 99%). The site
ofPnarbas A (90008000 B.C.) is the one exception, with
Figure 2 Mapof Anatolia with Central Anatolian Neolithic sites,
and chert sources discussed in paper (modified after Bezic, 2007;
Doherty,Milic, & Carter,2007). Archaeological sites: (1)
Catalhoyuk; (2) Pnarbas; (3) Can Hasan; (4) Mersin; (5) Askl Hoyuk;
(6) Musular.
342 Geoarchaeology: An International Journal 28 (2013) 340362
Copyright C 2013 Wiley Periodicals, Inc.
-
NAZAROFF ET AL. NEOLITHIC CHERT AT CATALHOYUK, TURKEY
a chipped stone assemblage consisting of 40% chert.Concerning
sources of chert raw material, Bezic dis-cusses the presence of
three source locales, only one ofwhich is located in Central
Anatolia. These are depositseast of Beysehir Golu, and two locales
in SoutheasternAnatoliaKahramanmaras and Sakcagoz (ibid, Map 1).In
spite of this, she is unable to provide much informa-tion regarding
actual raw material procurement.
Doherty et al. (2007) take us one step closer to un-derstanding
chert provenance through their case studyof the materials recovered
from the earliest levels ofCatalhoyuk. To Bezics list, they add the
ophiolite depositsin the Antalya and Mersin regions. Furthermore,
the au-thors present the results of visual and elemental analy-ses
of chert artifacts from the excavations at Catalhoyuk.They note the
presence of 23 visual chert groups, andfurther reduce these groups
to six or seven chert-typesbased on the formation processes at the
origin of eachmaterial. Environments with similar processes
reportedin geologic maps were used to hypothesize chert
pro-curement at ophiolitic deposits affiliated with the
TaurusMountains, and at volcanic areas north of Catalhoyuk.
While chert-sourcing programs in Central Anatolia andat
Catalhoyuk have generated thought-provoking discus-sions concerning
Neolithic chert procurement and use,none of these have, as of yet,
actually conducted anythorough raw material survey to test the many
hypothe-ses they have generated. Until now, provenance
inter-pretations have been primarily derived from
publishedreferences. Many issues may exist when only lookingto
geologic maps or reports in an attempt to discernsource locales, as
the interests of geologists and archae-ologists often differ.
Discrete chert deposits that would beof importance to
archaeologists may not find their wayinto the geologic literature.
Moreover, Central Anato-lia contains several environments where
chert may haveformed. Simply referring to those environments
closest toCatalhoyuk as possible procurement locales ignores
thecomplexities of prehistoric procurement strategies. Whilesuch
approaches may be an adequate first step in identi-fying potential
areas of chert deposition, pedestrian sur-vey is required to make
adequate statements of chert rawmaterial provenance.
GEOLOGIC SURVEY METHODOLOGYAND RESULTS
Survey Methodology
Chert artifacts studied at Catalhoyuk were made ofmaterials
formed from lacustrine, marine, deep ma-rine, radiolarian, and
volcanic-related chert (cf. Dohertyet al., 2007). Our survey
targeted geological environ-
ments likely to have produced these materials. A reviewof
geologic reports and maps (e.g., Akyol et al., 1963;Waldron, 1984;
Robertson & Ustaomer, 2009) facilitatedthis process. Chert
formation often involves the additionof silica into carbonated
environments. This silica mayoriginate from siliceous oozes formed
from the collectionof organisms with silica-rich skeletal
materials. Alterna-tively, hydrothermal systems can introduce
silica fromvolcanism or the upper mantle into lake or oceanic
en-vironments, wherein the Si oxidizes into SiO2. So
calledsecondary cherts develop when silica replaces pre-existing
sediments, although some of the visual or min-eral components of
the parent material are maintained.This happens through a variety
of mechanisms, includ-ing silicification resulting from the
addition of silica fromnearby volcanism (Lowe, 1999:8592). As a
result, whenlooking at the geologic history of Central Anatolia,
ourattention was drawn to locations where microfauna mayhave
increased silica solution in marine environments,and to locales
where hydrothermal systems could haveacted as sources of silica for
the formation of chert in shal-lowmarine environments, or for the
alteration of primarysediments.
Geologically, Anatolia is divided into three primarytectonic
unitsthe Pontides, the Anatolides-Taurides,and the Arabian
Platformcurrently separated by su-tures marking the separation of
these units prior to theOligocene by the Neo-Tethys and
Paleo-Tethys oceans.(Robertson et al., 2004; Okay, 2008). These
oceans areessential for understanding the distribution of
radiolar-ian chert in Turkey. Ophiolites, the preserved relics
ofthe Tethys oceans, generally represent former oceanicaccretionary
complexes (Okay, 2008:23). These forma-tions often contain
radiolarian cherta highly siliceousmaterial mainly formed by the
alteration of radiolar-ian oozes (Calvert, 1971), which themselves
are theresult of oceanic silica-secreting organisms
(Luedtke,1992:23)which were utilized by the Neolithic occu-pants of
Catalhoyuk (Bezic, 2007). During the mid-Cretaceous, the
emplacement of ophiolitic melange andoverlying ophiolites over the
Anatolide-Tauride terrainresulted in an expansive distribution of
these materials.As a result, ophiolitic melanges are a frequent
occurrencethroughout much of Anatolia. It is therefore
understand-able that previous projects have looked at
ophiolite-richregions (Antalya, Mersin, and Beysehir) as possible
lo-cations for prehistoric raw material procurement (e.g.,Doherty
et al., 2007).
Central Anatolia is demarcated by the Taurus Moun-tains to the
south, to the west by the Lake Dis-trict beyond Egirdir, the hills
north of the HaymanaPlateau, and to the east by the eastern edge
ofCappadocia. The region encompasses several sections
Geoarchaeology: An International Journal 28 (2013) 340362
Copyright C 2013 Wiley Periodicals, Inc. 343
-
NEOLITHIC CHERT AT CATALHOYUK, TURKEY NAZAROFF ET AL.
of the Menderes-Tauride Block, including the Tavsanland Afyon
Zones, and the eastern portions of the Ly-cian nappes (Bozkurt et
al., 1997; Okay, 2008). Dur-ing the Paleocene, continental
collision initiated the di-vision of the Anatolide-Tauride terrain,
which resultedin a south to southeast thrust pile. The lower
sectionsof this thrust pile were regionally metamorphosed in
thenorth, while the upper portions in the south formed largecover
nappes. There exists a subdivision of the Anatolide-Taurides based
on different metamorphic features, thethree primary zones include
the Tavsanl Zone in thenorth, a Cretaceous blueschist belt, the
Afyon Zone oflower grade high-pressure metamorphism in the cen-ter
of the Anatolide-Taurides, and finally the MenderesMassif in the
south, a region comprised of deformeduppermost
Cretaceous-Palaeocene flysch with Triassic toCretaceous limestone.
In spite of such variability, theseregions all have in common a
late Precambrian crystallinebasement, followed by a mixed
clastic-carbonated Paleo-zoic succession, and an Upper Triassic to
Upper Creta-ceous carbonated sequence (Okay, 2008:30). The
regionhas been the site of extensive carbonate platforms duringthe
Mesozoic, resulting in the deposition of shallow ma-rine carbonates
several thousand meters thick (Yalcn &Ciftci, 2002).
It was decided that two general regions held great po-tential
for locating raw materials of archaeological in-terest. First was
the western edge of the Konya Basin(Figure 1). Abundant ophiolite
deposits in this region in-dicated the possible presence of
radiolarian cherts. Sec-ond was the region surrounding the modern
city ofAfyon south of Eskisehir (including the Sultan Daglar,the
Sandkl Dag, the Karakus Daglar, and the YazlkayaPlateau). It was
believed that local volcanism mighthave promoted the silicification
of pre-existing sedi-ments, resulting in the formation of secondary
chert de-posits. Note that, for the sources reported below, wegroup
Suhut Chert and Suhut South Chert into a sin-gle source (see above;
Table II), and similarly considerHatip Chert and the Carsamba
Gravels as a single chertsource.
The Western Konya Basin
Survey of the western Konya Basin (Ylmaz, 2010)focused around
the modern villages of Sille, Szma,and Hatip, the western and
southern sections of theObruk and Cihanbeyli Plateaus, and deposits
affiliatedwith the Carsamba River which enters the Basin fromthe
southwest. Three chert raw material sources ofpossible
archaeological importance were located: (1)deposits of chert around
Gedik Tepe north of Konya,(2) ophiolite deposits adjacent to the
modern village
of Hatip, and (3) radiolarian cherts located along themodern
Carsamba River. Only the latter two producedevidence of knappable
material of sufficient size for likelystone tool production.
Although we use the term KonyaRadiolarite as an umbrella-term for
all radiolarites foundwithin the Basin and its immediate vicinity,
we havefurther divided this group to account for differencesbetween
Hatip Chert and the Carsamba Gravels. Whilethese materials broadly
relate to ophiolite depositsdiscussed in geologic literature (cf.
Okay, 2008), a morefine-grained approach is required to better
understandhumanlandscape interactions.
The western Konya Basin contains sections ofthe Afyon-Bolkardag
zone (Robertson & Ustaomer,2009:114), a lower grade
high-pressure metamorphicbelt, which is related to the
Anatolide-Tauride block(Okay, 2008:30). The Afyon-Bolkardag zone
exhibits ageological sequence typical of the Tauride region,
includ-ing a mixed carbonate-clastic Paleozoic series that is
over-lain by Mesozoic marbles. The region shows a low-grademedium
to high-pressure metamorphism. This is charac-terized by a common
occurrence of carpholite and localsodic amphibole (Candan et al.,
2005). Ophiolites andan ophiolitic melange overlay these
metamorphic rocks(Okay, 2008). The Hatip Chert source reported
hereafine-grained radiolarian chertis most probably relatedto such
ophiolites.
The Boyalitepe Triassic-Lower Jurassic platformlimestonesa
series of oolitic rocks, algae and foraminif-erous mudstones, and
reddish, thin-layered limestonesare located further south within
the Basin. This sequenceis overlain by a condensed, three-section
sequence,comprised of ammonite-bearing Jurassic marl, whiteand red
colored radiolarite-bearing limestones of LowerCretaceous age, and
an Upper Cretaceous Globotruncana-bearing stilolitic limestone
interbedded with red-coloredsilica (Valor & Tunay, 1996:1618).
On a more generallevel, present are the remains of oceanic
accretionarycomplexes, now ophiolitic melanges, which are
com-prised in part of radiolarian cherts of Triassic, Jurassic,and
Cretaceous ages (Tekin, 2002). Both geologists andarchaeologists
have suggested that Jurassic sedimentshold some of the greatest
potential for containing artifact-quality cherts (Hein &
Parrish, 1987; Duke & Steele,2010).
Regionally, the Konya Basin appears to contain vari-able types
of raw materials similar to those recovered atnearby Neolithic
sites, including not only chert but schist,diabase, and andesite
used in the production of groundstone objects (Baysal, 2004, 2009).
Previous projects,such as the Konya Plains Survey (KOPAL) (Baird,
2002),have visited much of the region and reported on geo-logic and
archaeological materials. Recent paleobotany
344 Geoarchaeology: An International Journal 28 (2013) 340362
Copyright C 2013 Wiley Periodicals, Inc.
-
NAZAROFF ET AL. NEOLITHIC CHERT AT CATALHOYUK, TURKEY
Table
IIDescriptiv
einform
ationpertainingto
chertsou
rces
discu
ssed
intext.
Subgrou
pDistin
ctiveVisua
lGen
eralDescriptio
nof
Dep
osit
Source
Nam
eUTM
Elevation(m
)Color
Ran
ge(M
unsell)
Cha
racteristic
sNod
uleSize(s)
andLo
calG
eology
Suhu
tChe
rtSu
hutC
hert
0290
007
4278
595
1312
5R3/4,4/6;5RP4/2,6/2;5
YR5/6,
6/6,
8/1;
5Y8/1;
10R4/2,
4/6;,6
/2,8
/2;
10YR
6/6,
8/2;
N6-N9
Materialhas
avery
waxyluster,
oftenwith
white
spottin
gor
darkgray
cracks,thou
gh
neith
erof
theseinflu
ences
breakag
epatterns.
Blockscanexceed
5m
in
diameter,w
hile
disartic
ulated
nodules
canha
veadiameter
of
upto
47cm
.
Che
rtoc
curs
aslargeblocksinthestream
exposurewesto
fthe
mainroad
andno
rthof
thesideroad
.Blocksareov
erlainbycarbon
ate
sedim
ents,and
occu
rad
jacent
tovariou
s
volcan
icmaterials(and
esite
,dacite
).
Suhu
tChe
rt
South
0284
511
4261
709
1178
5YR
4/6,5/6,7/4;10
YR9/2
Similarto
theprimarySu
hut
dep
osits.
13cm
diameter
in
cong
lomerate,pluson
e
0.5m
disartic
ulated
noduleinfield.
Aseco
ndarydep
osition
ofch
ertn
odules,m
any
occu
rringinaco
nglomeratelocatedon
the
hillsidewesto
fthe
Afyon
-Suh
utroad
(sou
thof
Suhu
t),b
utothe
rno
dules
arepresent
inthe
nearbyfield.N
oduleden
sity
isextrem
elylow.
Thech
ertsho
wssimilarvisualch
aracteristicsto
Suhu
tChe
rt.The
surrou
ndingland
scap
eisa
lake
bed
with
older
volcan
ism
tothesouth.
This
volcan
ism
may
have
caused
such
seco
ndary
dep
osition
.
Akd
ereChe
rtN/A
0307
808
4347
578
1204
5YR
6/6,
8/1;
10R5/4,
7/4;
10YR
6/2;
tran
sluc
ent;
N7-N9
Tran
sluc
ento
rmilkywhite
materialw
ithahigh
freq
uenc
y
ofblack
flecking.
Asilkyluster,
somematerialsha
ving
blue,
pink,or
brownhu
esat
dep
th.
Diameter
ofno
dules
rang
e
from
5cm
20cm
,while
materialoccurring
as
veinsem
bed
ded
with
in
thesurrou
nding
limestone
scanha
vea
width
ofup
to20
cm.
Asource
ofno
dular
andveinch
ert,seen
in
tran
sitio
nfrom
lake-bed
limestone
sinseveral
places.Man
yno
dules
occu
rwith
inthe
limestone
matrix,an
dman
yof
thoseerod
ing
outm
aintainsomeof
thismaterialasa
cortex.Nod
uleden
sity
isextrem
elyhigh
,with
artifactq
ualitymaterialoccurring
throug
hout
thesurvey
region
(hillsidean
dne
arbystream
).
Kony
a
Rad
iolarite
HatipChe
rt04
4703
6
4180
094
1174
5R2/2,
4/2,
5/4;
5YR
3/2,
4/2,
4/4,
5/6;
10YR
7/6;
N3-
N5
Opaq
ueor
semitran
sluc
ent
oran
geco
lor,with
some
instan
cesof
minim
alblack
flecking.
Nod
ules
rang
efrom
a
diameter
of1cm
40
cm.
Assum
edto
also
beaffiliatedwith
Kony
a
Rad
iolarite,a
materialw
idelydispersed
throug
hout
theregion
.How
ever,m
ore
high
-qua
litymaterialoccursat
thislocale,
certainlycapab
leof
produc
ingartifactual
material.Sa
mplelocaleisad
jacent
toparen
t
ophiolite
s,aco
mmon
presenc
earou
ndthe
fringe
oftheKo
nyaBasin.
Carsamba
Gravels
0462
021
4142
969
1023
5G4/1;
5YR
3/2,
3/4,
4/4;
10R3/4;
N5-N7
Stan
dardredan
doran
geco
lor
forAna
tolianradiolarites,with
someinstan
cesof
gree
n.
Nod
ules
rang
efrom
a
diameter
ofless
than
1
cm7
cm.
Localesinclud
eroad
cutsad
jacent
tothemod
ern
course
oftheCarsambariver,someno
dules
retrievedfrom
theCarsambaitself.Nod
ule
den
sity
isge
nerally
high
eron
theban
ksof
the
Carsamba.Afewno
dules
have
bee
nreco
vered
from
grea
terdistanc
esaw
ayfrom
theriver,
mostp
robab
lyaffiliatedwith
thealluvialfan
derived
from
theCarsamba.
Geoarchaeology: An International Journal 28 (2013) 340362
Copyright C 2013 Wiley Periodicals, Inc. 345
-
NEOLITHIC CHERT AT CATALHOYUK, TURKEY NAZAROFF ET AL.
research at Catalhoyuk (Asouti & Hather, 2001; Asouti,2005;
Fairbairn et al., 2005) has provided evidence forthe use of various
environments in the Konya Basin,which contain chert sources. Other
researchers work-ing at Catalhoyuk have indicated that the
CarsambaRiver was also utilized during the Neolithic for a
varietyof resources (e.g., Baysal, 1998; Asouti, 2005;
Doherty,2008), and was one source of radiolarian chert used
atCatalhoyuk (e.g., Carter et al., 2005; Bezic, 2007). Lo-cally,
the deposition of radiolarite nodules is variable inquantity,
quality, and size. We must consider that prehis-toric procurement
strategies would have been impactedby this varied material
distribution.
The following deposits have been previously docu-mented in the
Konya Basin (Akyol et al., 1963). ExactUTM coordinates of sample
locales are reported inTable II.
Hatip chert
Affiliated with the abundant ophiolitic materials that oc-cur
throughout Central Anatolia, the Hatip Chert sourceconsists
principally of radiolarian cherts disjointed froman ophiolite
matrix. Unlike other source-locales whereradiolarian cherts are
embedded within a melange ma-trix, Hatip Chert is located directly
adjacent to its parentophiolites. The source-locale is a series of
hills directlywest/northwest of the village of Hatip, 13 km
southwestof Konya. Modern quarrying activity occurs below
oneoutcrop from which samples were collected. The mate-rial occurs
as a bright orange color that is almost translu-cent, and as an
opaque orange radiolarite. Several nod-ules contain a series of
small black flecks. While orangeradiolarian cherts are present at
other locales adjacentto the Konya Basin, the unique black flecking
in HatipChert may be used to distinguish it from similar
deposits.High-quality nodules, easily capable of producing
chippedstone objects, average 815 cm in diameter, althoughlarger
nodules were witnessed. Only the most minimalerosion was observed.
In general, fewmaterials were seendeposited in secondary contexts
outside their primary lo-cations. This has important implications
for modeling ma-terial procurement (Shackley, 2002). We can assume
thatthe procurement of such materials must have taken placeat or
nearby those deposits mapped by the project. Lithicreduction in the
form of small bifacial cores was wit-nessed at the source. This
suggests that prehistoric ex-ploitation of Hatip Chert did occur.
It is also worth not-ing that the source lies only 37 km
west/northwest ofCatalhoyuk, with no natural features to block the
lineof sight. The only other source with such close proxim-
ity is the Carsamba Gravels that are 35 km southwest
ofCatalhoyuk.
Carsamba gravels
The term Carsamba Gravels, as used here, refers toradiolarian
cherts located in secondary deposits south/southwest of the Konya
Plain. These materials have likelyeroded from deposits in the
Western Taurides, and thosedirectly south of Beysehir Golu (such as
the BeysehirHoyran Nappe) (Valor & Tunay, 1996), which
havesubsequently been incorporated into the system of theCarsamba
River. Although the Carsamba River was notsurveyed to its full
extent, various points along its coursewithin the Konya Basin were
visited. Deposits 20 km eastof the gravels were surveyed, and
materials similar tothe Carsamba Gravels were recovered in lesser
propor-tions. Samples were collected from the rivers alluvial
de-posits at the southwestern edge of the Konya Plain. Ad-ditional
samples were gathered along the banks of themodern river, and at
adjacent deposits. Two distinct typesof chert were present: nodular
red Konya Radiolarite,and a unique light gray nodular material with
a distinc-tive rippled pale brown cortex. Both types were vari-able
in their ability to produce artifact-quality material.Principally,
the radiolarian chert of the Carsamba Grav-els is pale red to red
in color, with higher quality or-ange chert sometimes present,
especially in the bed of themodern river. At some locales
high-quality material onlyappeared as nodules of unusable size. On
average, nod-ules ranged from less than 17 cm in diameter.
Althoughthe Carsamba River was extensively used during theNeolithic
(see above), no lithic reduction activities werewitnessed at any of
the locations surveyed.
The Afyon-Eskisehir Region
The Afyon and Eskisehir regions were surveyed after dis-cussions
with local geologists suggested the presence ofhigh-quality
chalcedony north of the city of Afyon. TheAfyon-Eskisehir region is
part of the Afyon-Bolkardagmetamorphic zone. Abundant dacitic and
andesitic vol-canism is present along the western margin of this
re-gion. Such volcanism may have acted as a source of silica,which
could promote the silicification of sediments, andthe formation of
secondary cherts. Two sources of possi-ble archaeological
significance were located in the region:(1) high-quality chert
blocks near the modern village ofSuhut, and (2) nodular and vein
chert adjacent to themodern village of Akdere. These sources have
been pre-viously mapped by Akyol et al. (1963).
346 Geoarchaeology: An International Journal 28 (2013) 340362
Copyright C 2013 Wiley Periodicals, Inc.
-
NAZAROFF ET AL. NEOLITHIC CHERT AT CATALHOYUK, TURKEY
Survey results indicate that the Afyon-Eskisehir re-gion offers
an abundance of high-quality chert material,a perfect medium for
the production of chipped stonetools. In addition to abundant
siliceous resources, theoccurrence of volcanic materials that may
have beenutilized for the production of ground stone
implementsmakes the region all the more attractive
archaeologically(Baysal, 2004). Limited evidence of erosion helps
to de-marcate the region into discrete source areas. The ab-sence
of chert witnessed in secondary contexts indicatesthat material
procurement would have been localized atthe primary deposits. The
region appears easily acces-sible from the Konya Plain via a series
of watershedsand valleys that cut through the surrounding
mountains.Thissen (2002) has suggested that the Neolithic
occupantsof Catalhoyuk interacted with populations in this
region.While little evidence from current archaeological
excava-tions exists in support of this hypothesis, the presence ofa
single piece of meerschaum at Catalhoyuk may indicatesome
connection with the region surrounding Eskisehir,as this area
contains the only known source meerschaumin the region (Baysal,
2004).
Suhut chert
The principle Suhut Chert deposit contained the largestblocks of
artifact-quality chert found during the survey.This deposit occurs
in a streambed roughly 20 km southof Afyon and 12 km north of Suhut
along the Afyon-Suhut north-south road. The source is surrounded
byabundant volcanic materials (andesite, dacite, and vol-canic
tuff) overlain by a later carbonated component. Itis likely that
the chert is a silicified form of the surround-ing volcanic
deposits. The remarkably high zirconium (Zr)concentration within
this material may further indicatevolcanic origins (Lowe, 1999).
While macroscopic char-acteristics of this source are variable in
relation to color,the chert appears to have a very waxy sheen
unknownat the other sources discussed in this report. Color
variesbetween a pure and opaque white, to a pale pink,
brightorange, purple-red, and deep red. A common and distinc-tive
visual characteristic of the red and purple-red ma-terial is the
presence of crystal-shaped white inclusions.These may be silicified
phenocrysts present in the parentvolcanic material, which, after
silicification, remained asvisual markers of the cherts previous
form. Many otherblocks also appeared to contain cracks of dark
color thatspread across the material surface. However, neither
thecrystal-shaped inclusions nor the visual cracks have anyeffect
on knappability. Indeed, Suhut Chert is as good araw material as
any for the production of stone tools. Theinclusion-free material
offers the perfect high-quality ma-terial for the production of
sharp and durable edges. The
Figure 3 Various chert blocks present at the Suhut Chert source
(A) largeblockofwhiteandpinkchert, (B) small boulderoforangechert,
(C) close-up
of (A). Photograph by Adam Joseph Nazaroff.
largest blocks present can reach a vertical height of over2 m,
and may have a diameter of more than 5 m (Fig-ure 3). Minimal
erosion has resulted in materials scat-tered across the nearby
slopes. No archaeological materi-als were witnessed at the
source.
Suhut south chert
The Suhut South Chert source is located less than 5 kmsouth of
the modern village of Suhut. The source is acces-sible via a dirt
road that veers west from the main roadout of Suhut. The source
lies north of this dirt road inan open field south of limestone and
conglomerate out-crops that contain small (
-
NEOLITHIC CHERT AT CATALHOYUK, TURKEY NAZAROFF ET AL.
Figure 4 View of the Akdere Chert source (A) and associated
materials(B). Photograph by Adam Joseph Nazaroff.
is assumed that this source is in fact the result of sec-ondary
deposition of the Suhut Chert nodules. In supportof this
hypothesis, we noted that the Suhut South sourcecontained chert of
similar visual characteristics to thatseen at the Suhut source.
This included either the crystal-like inclusions or uncanny sheen
indicative of SuhutChert.
Akdere chert
The Akdere Chert source is the most northerly of allsources
reported here, and by far contains the most abun-dant high-quality
chert raw material we witnessed. Al-though we discuss the Akdere
locale as a chert source,the material is most probably chalcedony.
The AkdereChert source occurs on the summit and flanks of a
hilldirectly north of the village of Akdere, and in a
nearbystreambed (Figure 4). Nestled in a highly metamorphicregion
containing abundant marble and schist materi-als are a series of
Upper Miocene and Lower Creta-ceous sediments, at the southeastern
edge of which islocated the Akdere Chert source. Unlike the nodular
ma-terial and large blocks found at the Suhut Chert source,
Akdere Chert appears principally as veins or small boul-ders
within a surrounding carbonated matrix. Only onthe flanks and in
the river can one find nodules of AkdereChert outside of its
matrix, averaging 1020 cm in size.Small nodules of red, orange, and
green radiolarian chertare also present, but few are of artifact
quality. Suchchert is abundant throughout much of Central
Anatolia(cf. Okay, 2008), and is thus, as of now, not consid-ered
part of the Akdere Chert source. Rather, AkdereChert occurs as a
range of opaque white, to milkywhite, and finally translucent
material, sometimes witha very faint blue, pink, or brown hue, and
most of-ten with characteristic black snowflake inclusions.Minimal
lithic reduction was witnessed across theoutcrop.
PROVENANCE METHODOLOGY ANDRESULTS
Locating and mapping the above chert sources is aworthwhile
endeavor as it demonstrates the presenceof artifact-quality
materials west and northwest of theKonya Basin in regions not
considered by previous prove-nance programs. Additionally, we have
confirmed thepresence of artifact-quality radiolarite on the
KonyaPlain. However, to provide evidence that these localeswere
exploited by the Neolithic occupants of Catalhoyuk,we have
undertaken an analysis of the visual character-istics and trace
element concentrations of source materi-als and cultural objects
from Catalhoyuk and have useda combination of these to provide
evidence for the useof these locales during the Neolithic. Using
visual char-acteristics to discern the different chert types
presentat Catalhoyuk offers an initial approximation of
whichsources were used at Catalhoyuk. This is accomplished
bymatching visual categories to corresponding samples col-lected
from the Suhut, Suhut South, Akdere, and KonyaRadiolarite sources
with similar visual attributes. Milneet al. (2009) demonstrate the
utility of visually char-acterizing chert materials, especially as
a first step inmaking qualitative sense of an otherwise diverse
assem-blage. Such approaches excel in regions where chert de-posits
are visually distinct from one another. This ap-pears to be the
case in Central Anatolia, where eachchert deposit thus far surveyed
appears visually distinctin relation to at least some of the IAA
conducted. Re-gardless, a combined approach of visual,
mineralogi-cal, and geochemical characterization is often
preferable(Kendall, 2010). Accordingly, we have also used
EDXRFanalysis to further characterize both cultural and
sourcematerials.
348 Geoarchaeology: An International Journal 28 (2013) 340362
Copyright C 2013 Wiley Periodicals, Inc.
-
NAZAROFF ET AL. NEOLITHIC CHERT AT CATALHOYUK, TURKEY
Chert sourcing programs have used a wide arrayof techniques to
characterize material deposits. Amongthese, instrumental NAA (INAA)
is often preferred inchert sourcing (e.g., Glascock, 2004; Kendall,
2010;Huckell et al., 2011) due in part to a greater preci-sion for
several elements when compared to other tech-niques. However,
because INAA necessitates a nuclearreactor, choice of laboratories
for sample submission islimited, and analysis can be relatively
time consumingwhen compared to XRF. Other techniques employed
in-clude XRD (Malyk-Selivanova et al., 1998), inductivelycoupled
plasma mass spectroscopy (Evans et al., 2010;Olofsson &
Rodushkin, 2011), electron microprobe anal-ysis (Trogdon, 2006),
thin-section petrography (Bustilloet al., 2009; Rodriguez-Tovar et
al., 2010), visible/near-infrared (VNIR) reflectance spectroscopy
(Hubbard et al.,2004; Parish, 2011), and chert luminescence
properties(Akridge & Benoit, 2001). Our decision to use
nonde-structive XRF analysis was based on instrument acces-sibility
and future research objectives. Because of cur-rent legal
constraints within Turkey, destructive analysisof cultural objects
is often not an option. It is also diffi-cult to export samples
from the country. At Catalhoyuk,it can even be problematic to
remove artifacts from thesite. Because several geological and
archaeological stud-ies that seek to discern the bulk chemistry of
cherts havesuccessfully used XRF instrumental analysis (e.g.,
Girtyet al., 1996; Van Kranendonk & Pirajno, 2004; Olivareset
al., 2009), we determined that XRF was an apt choicefor
nondestructive analysis.
Sampling Procedures
Cultural objects were sampled from levels GP of theSouth Area
excavations at Catalhoyuk (Table I) due tothe apparent greater
diversity of chert material typeswhen compared with later levels.
Such a sample there-fore holds the greatest potential for capturing
the vari-ety of chert sources used at Catalhoyuk. In all, 686
ar-tifacts from the South Area excavations at Catalhoyukwere
macroscopically analyzed with IAA and sorted into22 possible
chert-types (not including debitage recoveredfrom flotation) (Table
III). Of these, 28 artifacts visuallymatching the known chert
sources were sampled for geo-chemical analysis. From the raw
material sources, we vi-sually analyzed 96 samples: 24 from the
Suhut source, 4from Suhut South, 30 from the Akdere source, and
38Konya Radiolarite from Hatip and the Carsamba Gravels.A subsample
(n = 19) was taken for geochemical anal-ysis. It was insured that
each sample analyzed was largeenough to produce reliable results
following Davis et al.(1998).
Visual Analysis
IAA included measures of color using the 2009 Geo-logical
Society of America Geological Rock-Color Chart,opacity, luster,
texture, fracture properties, and inclu-sions. Terminology for
opacity, luster, and texture wasadopted from Luedtke (1992:6870).
When samples con-tained multiple colors, each was recorded with its
relativeproportion and character of distribution (e.g.,
banded,mottled, etc.). Opacity, luster, texture, fracture
proper-ties, and inclusions were variable across both source
andartifact assemblages. As discussed below, IAA producesvisual
categories that comprises multiple macroscopic at-tributes.
Although many categories may share in com-mon one or several
attributes, the unique combinationsof several characteristics
constitute each visual class. Allmeasurements were taken with the
use of hand lenses(1020).
EDXRF Analysis
Geochemical analysis was conducted using a Ther-moFisher
Scientific QuantX EDXRF instrument for bulkchemical analysis. All
samples were analyzed whole byM. Steven Shackley at the
Archaeological XRF Labora-tory, Albuquerque, New Mexico. Sample
placement in-sured that the largest amount of X-rays possible
bom-barded each sample. Analysis was conducted for majoroxides
aluminum (Al2O3), silica (SiO2), and iron (Fe2O3),and trace
elements titanium (Ti), manganese (Mn), iron(as FeT), copper (Cu),
zinc, (Zn), rubidium (Rb), stron-tium (Sr), yttrium (Y), zirconium
(Zr), and thorium (Th).Additional information on procedures and
instrumentspecifications is available at
http://www.swxrflab.net.
When choosing a suite of elements for provenance as-sessments,
it is important to consider the possible ori-gins and diagenetic
processes affecting each element.During chert formation and
diagenesis, inclusions incor-porated from the surrounding
environment (e.g., clays,carbonates, iron oxides, or organic
matter) may altertrace element geochemistry (Luedtke, 1992:38;
Murray,1994:214). This process is often important for prove-nance
research, as inclusions unique to each locale maybe used to
fingerprint particular deposits. Murray (1994)used multiple
combinations of trace elements and ele-mental ratios to
discriminate among chert deposits thatformed under different
environmental conditions. Chert-sourcing programs often utilize
these geologic histories,manifest in the mineralogy and chemistry
of chert, tomake provenance assessments (Malyk-Selivanova et
al.,1998; Lowe, 1999; Doherty et al., 2007; Milne et al.,2009;
Parish, 2011). Mineral and chemical indicators pro-vide information
on characteristics unique to a materials
Geoarchaeology: An International Journal 28 (2013) 340362
Copyright C 2013 Wiley Periodicals, Inc. 349
-
NEOLITHIC CHERT AT CATALHOYUK, TURKEY NAZAROFF ET AL.
Table III Visual attributes of 22 chert types discerned from the
Catalhoyuk assemblage, South Area excavations levels GP. Note that
groups 20, 21, and
22 have probable correlations with the Akdere, Konya
Radiolarite, and Suhut chert groups.
Chert Possible Geochemical Primary Color Number in Artifact
Group Group Range (Munsell) Visual Characteristics Subsample
1 N/A 5Y 5/66/6; 10 YR 4/26/2 Sesame-colored brown chert with a
greasy luster A white
siliceous cortex penetrates into some samples.
n= 89
2 N/A 5 YR 3/24/2; 10 YR 4/2 Translucent dark brown material
with a disctinctive white
cortex that does not penetrate into the material (sensu
Group 1).
n= 63
3 N/A 5 YR 4/1; N6, N7 Fine-grained waxy gray chert, oftentimes
with a blue hue. n= 154 N/A 5 YR 2.5/23/2 Opaque fine-grained
medium-brown chert. n= 95 N/A N13; N67 Fine-grained waxy white
material, most often found with
burned exterior surface.
n= 7
6 N/A 5 YR 5/1; 5 Y 6/1 Coarse medium-gray chert with a greasy
luster and light-gray
speckling.
n= 25
7 N/A 5 BG 3/2, 5/2 Brilliant green or green-blue chert that has
a waxy luster and
a medium-coarse grained texture.
n= 1
8 N/A 5 YR 2/2; N4 Dark fine-grained pearly gray chert. n= 19
N/A 5 YR, 3/1, 3/2, 4/1; 10 YR 3/2 Dark green-brown chert, at times
mottled. n= 1510 N/A N1 True black fine-grained chert. n= 1011 N/A
5 YR, 4/4, 5/6; 10 YR 6/2 Dusty brown-red material with a grainy
appearance. n= 1312 N/A 5 Y 4/2; 10 YR 4/2, 7/2 Coarse brown and
yellow chert. n= 613 N/A 10 YR 4/2 Light reddish-brown fine-grained
chert. Not radiolarite in
appearance.
n= 5
14 N/A 5 R 3/6 Grainy and coarse bright red chert. n= 115 N/A 10
YR 5/2, 6/2, 7/2 Coarse-grained brown chert with a dull appearance
and
small, dark inclusions of unknown origin.
n= 8
16 N/A 5 YR 3/4; 5 YR 2/2; 5 YR 2/1 Mottled yellow-brown chert.
n= 217 N/A 10 R 4/6 Translucent red chert. Not radiolarite in
appearance. n= 118 N/A 5 YR 4/4; N9 Chalky white material with
siliceous orange-brown
components.
n= 1
19 N/A 5 RP 2/2, 4/2 Fine-grained purple material. n= 220 Akdere
Chert 5 YR 4/1; 10 YR 6.5/2, 6/6; N8 White or translucent
chalcedonic material, with variable hues
(brown, yellow, pink, blue), and often black (Mg?)
inclusions.
n= 232
21 Konya Radiolarite 5 YR 5/6; 5 R 3/22/2; 10 YR
5/4, 6/6 7/6; 10 R 2/4
Variable radiolarite materials, consisting of orange, green,
red, and brown materials as evidenced on the surrounding
Central Anatolian landscape.
n= 117
22 Suhut Chert 5 YR 4/6; 10 YR 7/2; 10 R 4/4 Very waxy chert
with a bright sheen, principally white,
burgundy, or orange.
n= 18
23 N/A Variable Samples were in various state of burning
(evidenced in the
presence of severe cracking, pot-lid fractures, or
discoloration), or weathering (determined through visual
patination).
n= 45
host sediment. This informs researchers of palaeoenvi-ronmental
processes, or the parent material (dacite, vol-canic ash,
sandstone, etc.) of secondary silica deposits.Analysis of more
mobile elements can also offer dataon localized environmental
conditions, diagenetic pro-cesses, or other local mineralogical and
chemical variabil-ity, which may leave an imprint on a particular
portionof a formation (Murray, 1994).
In this study, we use concentrations of Zr, aluminumoxide
(Al2O3), and iron oxide (Fe2O3) normalizedto silica (SiO2), to
discriminate among chert sources
(Table IV). Murray (1994:216) states that alongwith Ti and rare
earth elements (REEs), only Al andFe appear relatively unaffected
during diagensis. Hesuggests that additional elements, such as Zr,
mayalso remain relatively unaffected. We have thereforechosen this
set of elements because of their relativeimmobility and their
unlikeliness to be influenced bydiagenetic process (Murray, 1994;
Lowe, 1999). Thebulk chemical analysis performed by XRF does
notallow for the identification of the specific inclusionsthat
account for the presence of a given element.
350 Geoarchaeology: An International Journal 28 (2013) 340362
Copyright C 2013 Wiley Periodicals, Inc.
-
NAZAROFF ET AL. NEOLITHIC CHERT AT CATALHOYUK, TURKEY
Table
IVGeo
chem
istryforsou
rcean
dartifactm
aterialsan
alyzed
inthisstud
y,with
probab
lege
oche
micalan
dvisualprovena
nceassign
men
tsfora
llartifactm
aterials.