Вы находитесь на странице: 1из 174

4. 42 is frolll G. l-rtson. M. Wil'lm, amI Il 1 Wil'ol1.

l-'ossi1 bison ano Hlif"l'!s Irum un crly A!till1l'II1l;i1I",r"d

arroyo trap in 1117(,,4J . ( 1); !ir 17
Repradul,cd hy pCllllb\lon of thc S"dety for A.1I1<'1 han
Extnll"( 011 p. 2\1.1 r- fWIII 1\. J. Su!difr,', SI'''!II'd hY;l<'11;I"
Crushcr , gnawr r, and <"<llll'e1'" of Oll11l's. Na/w'{',
1970,227:1111 R'"JlTinlcd hy pcnuission f'romNI//u!"e
Copyril(111 1970 M"llllill,1ll .11\\111I,lIs I imin-d
COI'YMI(;IIT 1 'JK 1, IlY A( "111 MI( I'Ml ss, INI
AH. RIUII"I! 1\1!1 )(VJ 11.
NO l'ART 01' TllIS l'lIIllICAlION MAY tll. JU l'IIOHI'1 11> ""
O'" 1'.'1u.,11"1'11('1.1, INn \IIllN( 1'l1(I"!(OCOI'Y, MI nl1lllll\:I . 01< ",,"'\
litORA(,! ANlllIIllIIIVAI. M. WIIII<l111
IN WIlI'l IN(; "'IOM "] m 1'1I1i11SIH l(
Orlando, Florida 32881
UltedKin,do," Edillon pllbli,lIld l1y
:una ()Ya! Roacl, Loadon NWI mx
L1braryof Conlrus Cttlolnl in Publicadon Dat
Binford, Lewls Roberts, ata.
Bones: anclent roen and modern myths.
(Studles In uehaeology)
Blblloguphy: p.
tncruees indexo
1. Human sketetcn. 2. Animal remstns (Arehoteology)
3. Anthropometry. 4. Fcrensc osteotcgv. S. eones.
1. Tille. 11. Series.
GN70.B53 930.10285 811776
ISBN 0-12-100035-4
ISBN 0-12-100036-2 (Paperbaek)
'R "'TH> 1'" LNI n.I" r ...n, ", .....,.RI<...
'7 "' ,q
To the memory of David L. Clarke
It is the duty 01fheorefical hypothesis lo oulrun foct
so Ihal speculcuon o'erlecps Ihe present information
stcte ond pomts the woy,lhen careful accumulotion
01 lested data will revise the vchdity of the
theoretioo! posltton, which moy then lecp ohead
-CLARKE 11912'239)
- - - ~ -
Foreword x
Preface xv
AcknowJedgments xv
List of Figures and TabJes xxi
Abbreviations xxvii
Part 1
Documenting a
Iong-term archaeologicaI problem:
Concepts and observations 1
Chapter 1
Relics lo artifacls and monumenls lo assemblages:
Changing conceplual frameworks 3
The Relic and Monument Phase 4
The Artifacl and Assemblege Phase 6
Chapter 2
Middle-range research and Ihe role of actualistic sludies 21
The Paradigm-c-One's Cuide lo Describlng the World 23
Theory-One's Guide lo Explaining the World 25
. ~
Part JI
Middle-range research-
In search of methodology
Chapter 3
Pallerns of bone modifications
produced by nonbuman agents
Previous Approaches lo Understendtng the Significance
of Broken and Modified Bone
Skeletel Disarticuiation
Tooth Merks
Other sources of Surficial Modificalions un Bone
Bone Breakage and Oestruetion by Animals
Modifications by Analomical Part
Breakage Unrelatecl to Consumption: Trampling and Bone
Bone Modtcatons end Methodology
Chapter 5
Assemblage composition: Patterns of association
slemming from the behavior of man
versus that of beasl
New World Beginnings-Man as the Agent
Artcen Problems end Assemblage Composltlon
Other Types of Assemblage Variability
Studies of Assemblege Compositon Ceused by Beests
Observenons of Wolves and Therr Behevor
Information on Latr Hehavior
tnformetton on Kili Behavinr and Comparsons
Control Collections of Animal-Structured Assemhlages
Part III
Putting our knowledge to work:
Seeking to know the past
Chapter 4
Human modes of bone medication
Previous Approeches tu Understanding Broken and Modified Bone
The Control Dala
Dtsmemberng Strategy
Butchering Marks
Cut Marks: Their Forro and Placemenl 00 Specific Bones
Chopptng. Bone Breakage and Butchering Techniques
Gnawing by Humans
Martow-Bone Breakage
Control Colleetions
Breakaga Related in Other Forms of Bone Processtng
Man-to-Man Compansons. or Alternative Human Behavtors
Comparing Man and Beest
Morphology of Borre Breakege
Chapter 6
Applicalion: A new look al Olduvai Gorge
Olduva Gorge, a Challenge to Our Methods
Analysis of Olduvai Fauna
Bvaluang rhe Degree of Destruction Suffered by the Olduva
Assemblage Composltton-cOlduva Gorge
Chapter 7
General conclusons
Lewis Binford is probebly the mosl influential
and sttmulatve ercheeologtst in recen! years. in
respeel lo hs insistence on hypothesis testlng and
theory building. Thie book is a continuetcn of tha
best of his previous methodological work and will,
wthout doubt. once agaln redtrect archaeological
and paleoanthropological researches in importanl
and 115efu1 waya.
As always Binford has set B grand tesk for him-
selt. and here he has made a major postve efforl lo
edvence rnethod and theory in regard lo severa! as-
pects o human versus animal subsistence behevtors
and thetr attendent restdues. From ha observetons
Oh Alaskan Eskimo butchery and meat-processtng
procedures, end his analytical observertons on wolf
kills and ther letrs. he has recogntzed a series of
distinctive [vdiagnostic"] patterns uniquely com-
mon to each set of circumstances. This builds on the
earlier work of Binford and 8ertram and of Blnlord
(Nunomiut Ethnoorchoeology), but extends and
amplifies thal work very substantally. An ncrees-
ingly approprtete corpus of comperettve data s
beng eforded by ethnoarchaeologfoal studtes in Af-
rica [Icr example, by 1. vellen. D. Gifford. D. Crader),
by various tephonomtc sludies (for example. by A. K.
Behrensmeyer, A. nm. P, Shipman and assocatesj.
and by various studtee of camtvore (parlicularly
byaena den) accumulations (by A. Sutclffe. R. G.
Kien. J. A. Maguire, G. Hani and assoctates. for in-
alance]. A set of analytcal procedures has been de-
vsed Ihat seems to afford measures of a variety of
dfferences between (nonhuman) marnmal osteolog-
leal assemblages in regard to composilion of unitary
and fragmented body parte. and ther respective
condition and preservation.
Ultimately Binford is ccncemed "with the de-
velopment of dlagnostlc procedures for recognzing
or identifying the agents which might have been re-
sponsible for the faunal facts preserven in an ancent
depoet." 1 belteve he has made substantive method-
ologtcal contrtbutons approprtate toward resolution
of thts very complex set of problema. These nclude:
the recogntuon and differentiation of animal
(carnivore) and human procedures (behaviors) in
carcess dsmemberment and perttttoning
the eJuddation and descrlption of pettems of
bone breakage and explication of degrees of destruc-
tion ("ravaging") by camivores (particularly, but not
exclustvelv canids)
the thorough inventory and systematic descnp-
non of the disposition and cberecter ot traces on
bones resultng from human-skinning, butchering,
and meat-processing practices
the development of analytical procedures (mea-
sures for ascertainng the [expected versus actual]
compostton of various borre assemblages (drawn
from assembleges of known derivalion) through Ihe
employment oC severa) dtegnosttc ndices of the fre-
quencies of body perts, complete or in various stages
o fragmentationldestruction
Binford has esserted. largely justifiably, thet
have been un-
duly caveler in ther treatment of bcne assembleges
In general, and conjont arlifact-bone assemblages in
particular. He insists thet "cleme for the use of lools
should be supported by Ihe cteton ef marks pro-
duced by 100Is." Although there are sorne notable
exceptions in erchaeologcal OCCUITences of late
Plestocene ege {fcr exarnple, Plncevent}, and even
in Mousterian occupations Ifor example, La Quina).
lt ls not unreasonable for Binford to teke this stence.
One can only applaud the vlew thet progress is nol
"rnede through diseoveries. . whieh are treated as
self-evident In thelr meanlng," and thal "basie re-
search makes possible Ihe reliable asslgnment of
meaning to observations." lt ilO agaio lo Ihe polot lo
insist Ihal Ihe developmenl aod telOting of "mlddle
range Iheory" is critical lo Ihe realization of a scien-
tific sludy of Ihe human eareer. It ls, unfortunalely,
also true Ihat Ibere is substantial "mylhology" in
paleoanthropology, even "modern" paleoanlhropol-
ogy. Hopefully. Binford's posilion and efforts will
ultimalely provoke others to develop and advance
melhodology appropriale lo Ihe resolution of what
are clearly real problems in a proper understanding
of Ihe pasL No one familiar with Ihe slate of Ihe art
would deny Ihat there is sliIJ mueh to be leamed
when (and 10 there is subslantive progress in the
development of melhodology and analytical proee-
Binford has soughl to analyze Olduvai faunal as-
semblages, mostly from reputed hominid oecupation
places, on the basis of published accounls provided
by Mary Leakey In Olduvai Gorge, Volume 3 (1971).
His analyses are bold and imaginative, and the resull
provocative In the exlreme. However, are his lOweep-
ing conclusions justifiable and realistic? There are
subslantive problems concerning Ihis corpus of
faunal material; hence Ihere are problems concern-
ing Binford's efforts al analysis, most of which he
reeognlzes. Herelofore Ihese alOsemblages of animal
resldues have no! been studied and inventoried sys-
tematically for breekege pattems. for survivorship of
skeletal parts per individual, oceven collectively for
a particular texon: and "cut marks" as evidence for
butchery-dsmemberment have nol been reoorded.
Maps of spalial dtstrtbuttons of such bone-ertact
concentratons have been publshed. but the record-
ing of particular laxa and body parts do no! enable a
linkage lo be effecled betwcen theso dislribulions
and Ihe faunal inventaries.
The difficulty (and hence limitations) of the
faunal sumrnary in the Olduvo Gorge, Volume 3. is
thal the overall frequencies of body parts for large
mammal speces are summarized (p. 276), bu! there
is no full breakdown of body perta, or an indication
of Iheir preservallon for individual laxa. (However,
A. W. and A. Genlry [GeoJogy, Bulletin of the British
Museum of Naturol History, 29(4) and 30(1), 1978]
have provided relatvely fuIl, but stil l ncomplete
body part dala Forthe Bovidae, and these dala often
do nol malch thal tabulaled in Olduvai Gorga, Vol-
ume 3.) Moreover Ihe limb side is never staled, and
those dala are critical lo the realilOtic calculation of
MNl values.
Mary Leakey early recognized, and on several oc-
casions has stressed, the several sedimenlary envi-
ronments wilhln which Ihese various occurrences
are situated, and has also directed attenlion lo Iheir
slate of inlaetness or exlenl of dlsturbance. Thus, she
distinguished "living f1oors" {Iargely inlacl hominid
occupatioo occurrences), possible {hominid) kili
sites, channel fills, and "diffuse" occurrences. Bin-
ford's analyses have been direcled loward Ihe eluci-
dallon of Ihe nature and exlent to which a diversily
of nalural as opposed to (or in conjunction with)
hominid faclors may have been responslble for Ihese
accumulations. He has been concerned wllh discero-
ing events and formation processes, and Ihus In Ihe
degree of resolution afforded by occurrences rela
tive, In part, lo Ihelr inlegrity. Binford condudes Ihal
a diversity of agencies were probably responsible for
Ihe slale of Ihese occurrences. He has disceroed
ous pattems and degrees of bone destruction (and
preservatlon), and employing faclor analysls has
soughl fo clemonslrate resemblances with modero
analogues, and thereby lo infer the role of various
agencies in producing Ihe various accumulations.
Binford eondudes Ihal "Ihe large. highly pub-
llcized sites as currently analyzed carry HUlespecific
intormation about hominid behavor." He belteves
thal "arguments about base carnps, hominid hunt-
ing, sharing of food, etc. are cerlainly premature and
most likely wildly naccurate." This pretty strong
statement ts probably argueble. However. he does
accepl as demonstreted "homtnld scavenging 10fJ
Ihe kills and death sttes of other predator scavengers
for abandoned analomical parts of low food uulty.
primarily for purpnses of extrecttng bone marrow."
Thls conclusion, if further substantated, is itse!f of
mejor signifieanee for understanding Ihe dielary and
behavoral adjuslmenls of Plin-Plestocene homi-
nids. Recenl studes of bone assembleges, with arti-
facl assocatons. from Koobl Fora (by H. Bunn) and
fram Olduvai Gorge (by R. PolIs and H. Bunn) now
promise lo document in detal the nalure and extent
of hominid butchery practices evidenced al these
earHer Pleistocene sltes. These researches also do
confirm the complexily of formalion processes for
particular hominid "occupation situations," as Bin-
foed has concluded.
There has been, and is In progress, a variely of
research Ihat should provlde substantive dala rele-
vanl lo Ihe resolulion of many issues abou! which
Binford has expressed his profound skeplicism. Sev-
eral monographs (under Ihe editorshlp of J.-K. Woo)
on Ihe Horno erectus locality ofChoukoulien 1are in
preparation and should afford, al lasl, relevanl de-
lailed dala on those bone assemblages, and-
hopefully-the exlent lo which Ihey reflecl activities
of carnivores, early humans, or olher agencies, or a
combinalion of many facloes, al Ihal importanl site.
Claims have already been made (by P. Shipman and
associales) thal a distlnclive Theropilhecus
owaJdi-rich occurrence allhe Olorgesailie localily
reflecls an unusual form of hominid predalion (de-
spile Ihe apparenl absence of butr.hery Iraces!) ralher
Ihan sorne olher natural calaslraphe. A delailed
study of Ihls and olher bone assemblages is currenlly
being underlaken by Chrislopher Koch.
Binford offers sorne pilhy commenls relativc lo
our own work al Ihe Acheulian localilies of Torra Iba
and Ambrona in central Spain. It s worth mention-
ing that when those sttes were nittally excavaled
(1961-1963), initially under my direction and sub-
sequently joinlly with L. G. Freeman, no one was
concemed with bone assemblages in any of the ways
which have become more commonplace 10-15 years
hence (in facl the Leakey's became concerned with
such matters only in Ihe late 19605 as a consequence
of the extensve excaveton program et Olduvet
Gorge). Unforlunalely, and for severel complex rea-
sons. those faunal collectons from Ihe Spantsh sttes
were never fully analyzed, beyond texonomc iden-
tceton. and Ihus fracture patterns. cut marks. and
delails of dismembermenl and butcherv were only
superficially recorded in the fleld. These faunal col-
leotons are now relocaled and in course of detatled
study. Renewed excavation (in 1980) al Ihe Ambrona
locality has revealed an abundance of evdence rele-
vant lo butchery, dismembermenl, end bone break-
age, as well as human manipulalion and fashioning
of bone and ivory. Sedimentary processes at Ibis site
are more adequalely underslood than before, but the
silualion is an unusual one (for Europe) and Ihere is
slill much lo do.
There is. in fac!. a diveesity of research com-
pleled, and presumably in couese of pubHcation, or
now in progress, which addresses Ihe deficiencies in
paleoanlhropological sludies with whlch Binford is
so inlensely---and rightly so--concemed. These
range from the Upper Paleolithic of Ihe U.S.S.R. (and
weslern Europe), Ihe so-called "M.S.A." of easlero
and soulhern Africa, the Mouslerian ofmany parts of
Europe and weslern Asia, the AcheuHan (and lis
facies) of caves in soulhweslern Europe, and cave
and open-air occurrences in Europe which mayoit is
Ihought by sorne, testify lo Ihe Initial occupation of
Europe by hominlds. In each and all inslances Ihese
researches will gain subslantially from Ihe analytical
procedures developed in Lewis Binford's provoca
live and stimulating book.
Universily of Cali(ornia, Berkeley

This book developed from en eructe started in
)anuary 1979. I had made observations on wolf be-
havior while conducting ethnographc research
among the Nunamiut Esktmo of North Central
Alaska. 1 though that rny obeervetons might be o
sorne value to those engeged in taphonomic studes.
As 1seerched Ihe literature for sorne comparatve
material on assemblege composilion, 1 reelteed that
my data on denso bone hreakage, and psttems of at-
tnuon were untque. and Iha! there was more lo say
than ortgtnally planned. The peper wes then ex-
panded lo a short book. in which 1descrtbed butch-
ertng and butcherng rnarks. and competed marrow
cracking 85 performed by man and by animals. 1 ex-
pended my llterature coverage ID deal with Dert'e
erguments, read the Iascinattng lterarure treating
Choukoutten, and del ved into Ihe German luerature
that clearly foresbadowed the "osteodontckerauc"
argumenta of both Breuil and Dart. As 1 treated the
problema uf pseudotools. butchering petterns as
manifesl in bone modification, and Ceorge Fnson's
erguments abcut muscle stnopng. I also begen lo
envsion sume Interesting conctustons stemmtng
from my data and researcb. These conclusions were
nol related lo Ihe compostuon of wolf kills and dens.
which now seemed pcrtpheral lo the !hrust of the
"smflll tc(:hnical book" developed Ihus far. Then
Diane Gifford lenl me her copy of Andrew Hill's im-
portAn! doctoral fhesis, submilleci to the llninrsly
ofLondon lHilI1975). Hill's thests contained dala on
assemblage compostuon for animal kili and death
stes in African setttngs. Now 1had comparattve rne-
lerial nr viewing my dala in a broader context. I was
fascinated to dtecover Ihal ~ b l a g e ' ~
tion of remans al wolf km sites in nonh 'Cent1'al
Alaska was essentially tbe same as the composilion
of assemblages from animals of comparable SiZ6 in
Africa, where the predators were unknown endror
variable and the prey-centelopes and equids-so
very differenl from the certbou of Alaska. 1 began to
work up tha assemblage composition comparsons
between HiIl's dala and rny own, in Ihe process gatn-
ing a new eppreclatton for meny of the facts and
thetr lmpl catons
As l worked on the kili dala. 1 ncreestngly
wished for sorne oompereuve den material. lt was al
ths point thet Richard KIein arrived in Albuquerque
to deliver several lectures. He had prevously sent
me offprints of many of his articles treating Ieunel
assemblages from South Arrica. NaturalIy when he
was here we had sorne intense conversattons ebout
feunal assemblages. Richard responded to my moan-
ng and groanlng regardlng the lack of dala 00 ani-
mal denso He had dug one! He promised to send me
an offprint of his report on his relurn lo Chicago. The
day il arrived was full of excitement for me, since 1
had already formed sorne ideas as lo whal a den
should look [ike based on Ihe kili sife assemblages
and the bints 1 hed from my Alesken material. On
.Inspectng hts data. 1 realizad thet Ihe assemblage
had suffered heevy atlrilion, and Ihal what tbe sur-
viving bones gave me was mora a reecon of Ihe
reletve slrength of the anatomcel pert rether than a
clear pcture of tbe parls transponed by animals to
ther denso Could 1 reconstruct the original as-
semblage using the lechniques developed earlfer for
understandlng attntion? (See L. R. 8inford and J.
B. 8ertram 1977.) As wl1l be seen. 1 was successful.
and the correspondence belween the result and the
data from Ihe kills was remarkable. Wilh this test
comparative component in place there was a kind of
closure and a general methodology seen outlined.
Clearly what was needed. was an application! rchose
lo implement my approach on the importanl faunal
materials from Olduva Gorge. 1 had played around
earlier with sorne of the models developed In my
Nunamiut study (L. R. Blnford 197Bb]and found that
there eppeered to be sorne "fits" between my models
and the actual faunal data. My rnpresson at thet
time, however, was that all the variabillty in the
Olduvai matertals was probably referable lo
nonhormnd behevor. That s. the fauna was back-
ground natural lo geologtcel deposite In fr-ica, de-
riving from the conlinuous aclion of predators and
prey as well as natural deaths in soch a fauna-rich
1had completed the book up through the rst half
of the Olduva chapter when Jim HilI and Jan Orcutt
visited me. They were conducling multivariate
analysee of settlement data at Ihe University of New
Mexico computer center. Jan had an operatonal pro-
gram for carrying oul factor analyss. 1 pul together
the Olduvai dala and Jan made severa! runs ustng
different essumpttons. The fascinaling result. which
in my opinon did isolate paltened cnnsequences of
homlnd behevor. was totally unexpected. Con-
sequently Ihe wrlting of the last half of the Olduvai
chapter and the conclusions was sheer excilement,
rather than the normal labor of finishing a book
whose "outcome'' has been known Ior sorne time.
The writing of this book has been an inlellectual
adventure, with several importan! turning potnts. as
has been indicated. 1 arn convlnced that a basic
methodology has been outlined and that findings of
importance have resulted. The methodolcgy Is nol
"oleen." and much more tntormaton regarding both
kills and animal transport of fauna s needed, as is
much better conlrol on the anatomical facts of bone
densily or hardness so importan! to reconstructtng
assemblages Irorn survlvng elements. 1 hope that
olhers will be motivaled to join in the research
needed lo move from lhis outline to a robust set of
melhods for giving meaning lo the facts of the ar
chaeological record.
In many cases acknowledgments are a kind of
formal recitation of normal relationshps. in which
an author thanks the typist, the photographer, ac-
knowledges formar teechers for nsptreton, end
closes by acknowledging a spouse for putting up
with the lensions of writing. This format is germane
lo my situation, bul Ihere is in addition a compli.
caled set of intellectualand lechnical debts Ihat must
be acknowledged as exlraordinary in Ihe wriling of
Ihis book.
Wilh regard to my experience wilh wo!ves and
Ihe dala coJlecled, 1 owe a ver)' special debllo Iwo
men who continually prodded me iolo recognizing
thal sorne of my assumptions were perhaps nol war-
ranted in workinR with Ihe tundra fauna of Brooks
Range Alaska. One was Simon Paneack. who played
skeptic with me as 1 asked him queslions about the
fauna] materials my crews were refilularly re;overing
durmg the Beld seasons between 1969 and 1973. For
inslance, if I described a coileclion made or field
observalions recorded 00 fauna fram a parlicular
pla;e, Simon would smile and say, "How you knaw
it's nol lndians or wolves?" He said Ihal lo me so
many times Ihall began altempting lo find out what
wolves "Iooked like" when viewed from Ihe
perspective of fauna. 1tried lo do the same lhing for
"Indians" hui was nol successful. The second man
was lohnny Rulland, who in his quiel way pointed
out lo me my first examples of wolf kiJIs and enum-
erated all the charactenstics that led him to distn-
gush a wolfkill from ene madeby his fellow huntere.
While 1W85 working al Anaktuvuk. Boh Stephen-
son of the Alaskan Departmant of Fish and Game
began his extraordinary study of wolves and Ihe
Nunamiut knowledge of wolves. This stimulated
considerable interesl snd discussion of wolves
amoog the Nunamiul hunters with whom 1 was as-
sociated, and in turo I was the beneficiary. 1am par
ticularly indebted to David Mekiana, who beca me
in/erested largely through slimulalion fram lustice
Mekiana, who was working closely with 8ab
Slephenson. David went out af his way to nform me
of evidence of wolf behavior.
Although the collection of wolf data was largely
my project done In conjunction wlth elhnographic
work, two members of my crew, Dan and Allison
Witter, were unliring in Iheir nterest In fauna and
did the nasty ob of collecling bones fram one of the
majar dog yards al Anakluvuk. Dan developed the
first stages of the ciassification af destruction and
madification presented in Ihis book. He also oro
ganized and recorded the marrowcracking experi-
ment, which serve as imporlant contrals on the
tool-assisled breakage patterns described here. It is
one of my major regrAtf Ihat Dan could nol continue
working on thA Alaskan malerials wilh me, bul
academic consderetons as well as a [ob in Australia
led him in new directions befare this work on fauna
'ctmld be undertaken from the analycal perspectiva.
A number of people have been important in con-
ditioning the charecter of tbe final form of this man-
uscript. AH of the drafted graphs end charts were
drawn by Dana Anderson. who never failed lo pro-
duce the exacl iIlustration I had in minrl. The tllus-
trations of bones used in demonstretng the cut
marks (Figures 4.06, 4.11. 4.16, 4.20-4.22, 4.25-4.32.
4.36-4.39,4.48, and 4.53) were all drawn from actual
specimens by [udith Browne of Albuquerque. ludy
has made a meter contnbutton to the utility and
quelty of this bcok.
Mosl of the photographs of bone breakege and
modtcaton were laken by my daughter. Martha R.
Binford. and James Moore. These photogrephs are
crucial lo the types of comparative argumente that
maka up mueh of the dismemberment and breakage
chepters. Martha and Itm spenl long hours propptng
up scales. adjusting llghtng angles. building sup-
ports for specimens out of all kinds of Ihings so the
engle would be right or the alignment between
spectmen and seale would be un the same plane. For
this Investment and the fine pietures that resulted
(Figures 3.02, 3.10-3.16, 3.18. 3.20-3.30, 3.32, 3.33,
3.40-3.45, 3.51-3.53, 4.50, 4.56, 4.57), I am most
grateful. Afler Ihe initial "spree" of picture taking hy
Martha and lim. r kept diseovering otber subjects
that needed ilIuslralion and Martha took all the
additional pholographs unassisled {Figures 3.01,
3.03, 3.05-3.09. 3.17, 3.19, 3.31, 3.34-3.39, 3.46-
A number of people eontributed pholographs that
(hey had taken and I requested. This assistance is
acknowledged in the following list: Roberl Stephen-
son, E'igures 5.01, 5.03, 5.04, 5.11: James OConnell.
E'igure 4.43: Jean-Philippe Rigaud, E'igure 4,07.4.10,
4.15,4.16, and 4.41.
Through suggestions and actual research, several
pImple have helped me cover sorne of Ihe foreign
language Iileralure. Miehael Jochim responded lo
one o my cries for help by ~ u p p l y i n B copies of arti-
eles from Ihe German literature lha! were not avail-
able in Albuquerque. Furlher "id wilh the German
Iiterature was provided by Larry Todcl. a stuelent in
the anlhropology departmen!. In aelrlition, my friend
James Finley of the biology department here in AI-
buquerque also helped out with sorne translation
when my nwn sktlls failed. Iean-Philippe Rigaud
duplicaled Iwo Important manuscrtpts unevetlable
in Albuquerque, and euggested severa! otber refer-
ences. Lawrence Straus helped by solving the mys-
tery of severa] French idioma, and he graciously lent
me sorne Interestng end rare articles from the early
French Iiterature. F. Clerk Howell c!arified several
points for me regardtng Torralba and Ambrona and
sent me literatura as a further aid. James ludge
helped eonsiderably in tracking clown sorne of the
literature treating early man in the New World.
Intellectual slimulation and nspration during
the course of writing thts book carne basically from
long telephone cnnversations with Diane Gifford, in-
tense conversatton with Richard Klen. and general
dscusstone of the problem wlth Bob verra. Iohn
Pfeiffer, and Stanley South. To those good friends, 1
am sineerely grateful.
Severel typtsts were involved in the preparation
of ths menuscrtpt. Lisa Edelhoff agan treated my
rough draft with gentle and considrate skill. Fred-
die Height. Marilyn Daily, Ruth Stewart. and Louise
McGuff elso prepared major sections of the rnenu-
script. Their commitment to a good job s greatly
A rough draft of the manuscripl was critiqued by
six highly qualified readers: George C. Frison of Ihe
Universly of Wyoming, Diane Gifford of the Univer-
sily of California, Sanla Cruz. f. CJark Howell of the
University of California. Berkeley. Richard E. Morlan
of the Archaeological Survey of Canada, Timothy
While ofthe University of California, 8erkeley, and a
final unidentified reviewer obviously wilh interests
in the New World. These reviewers provided me
with a truly impressive bocly of criticism, thal made
it possible far me to clear up numerous vague or
misleading parts of the manuscript. Their responses
forced me to look agaio at the manner and substance
of my trealmenl of others' dala or ideas. In adelition. I
have added (Iargely in notes) numeraus volunteered
pieees of informa\ion that have strengthened the
presentation r:omdderably. I cannot thank lhese
readers enouRh. I hflve never had such a body of
Ihoughlful. r.onslruclive. ami candid r.ritil":ism. To Ihe
six who look it seriously anel djrl lneir hesl to find
tbe weaknesses and points that needad clarifica-
tton. l am very grateful.
Funding for the fieldwork during the course of
which the dala on Alaskan wolf behavor was ob-
rened carne from the National Science Foundaton
and the Wenner-Gren Foundation lar Anthropologl-
cal Research. To these agencies, 1am most grateful.
List of Figures and Tables
Figure 3.01. Punctures made by animal teeth. 45
Figure 3.0Z. Animal-produced pitting and punctures on
the distal metapodial. 45
Figun 3.03. Animal-produced extensively pitted
bone. 4fi
Figure 3.04. Animal-produced pillad and scored compael
bone surfece. 47
Figure 3.05. "Compressor" pttted and seored by animal
teeth. 47
Figure 3.06. Compect borre scored by enfmal teeth. 48
Figure 3.07. "Arte MobUlar" produced by acds in como
pect root rneeses un a shoop jaw from the Bear ste. Anak-
tuvuk Pass. 50
Figure 3.08. Detall or bone in Figure 3.07. 50
Fillure 3.00. "Cbenneted" beeekage ot bones by ant-
mela 52
Figure 3.10. Highly polished end of a gnawed borre. 52
t'igure 3.11. Pifiad and radially scarred bone, the result of
extensiva gnawing. 52
Fillure J.12. "Pressure flaking" on Ihe end of a gnawed
bone resulting Irom chipping heck the edge 5J
Figure J.1J. Extensively ehtpped-baek edRes of gnawed
bcne. 53
"'igure 3.14. Chlpped-back fldges shnwing mtcrodentlcu-
lated effect. 53
Figure 3.15. Externel Iace of chipped-haek bona showlng
oblique tooth scarnng 53
Figure 3.16. gxtarnal faee of chipped-back bane showing
looth scerrlng from "slipping clown" the bane wtth the
teeth. 53
Figure J.17. Examples of channeled and chipped-back
bones commonly confused with human workman-
shp. 54
Figure 3.18. Rounded edge of gnawed bone. 56
Figure J.19. Collepsed cylinders with dtsttnctive
chtpped-back ends. 56
Figure 3.20. Long "longttudnally" spllt fragmenta with
denculated ends. 57
Figure 3.21. Group of "ridge-crtls! removed" f1akestypical
or animal gnawng. 59
Fisure 3.22. Pseudotools produced by animals chewing al
rtdges. 59
Flsure 3.23. Delail of poin! on a pseudotool shown in Fig-
ure 3.22. 59
Fisure 3.Z4. Animel-gnawed cranium with attached
entlers Icaribou). 61
Figure 3.25. Antmal.gnawed paletea (caribou) showtng
dtsncttve channeling and crenuletlon ot thln bone. 61
FIsure 3.26. Crenret dtsc as eommonly produced by gnaw-
ing cands. 62
Figure 3.27. lnitlal deslructlon of sheep mandlbJes by bosh
dogs (Iower Iwo) and wolves (upper Iwo). 62
Figure 3.28. Sequence uf carbou mandibular deetructton
by both dogs (Iower two and wolves (uppar fourj. 63
Figure 3.Z9. Wolf mandible [Iower specimen) and dng
mandibles (upper specimenl dsstroyed by wolves. acrh
sper-imens were recovered frcm wolf denso 64
Figure 5.12. Relationship between Alaskan wolf den end
. 223
Fi"uhl5.13. Relafinnship between dscree end paltmpsast
wolf kili assembiegea. 223
Figure 5.14. Relalonship between eeconstructed Alaskan
wolf den and kili assemblagea. 228
FilUrtl 5.15. Relatlonship between reconstructed dscrete
end pallmpsest wolf kili essemblagea. 228
Flsun! 5.18. Camparison between medrum-tc-larga and
small ungulstes introduced lo antmal dens-Swartklip
reconslructed (aune. 228
FiSure 5.17. Compertson between medium-Io-large and
small ungulatss. parls remaining at killlocallons. 229
Figure 5.1B. Comperson between kills and dens for
medlum-to-Iarge anmels. 229
Fisu",5.19. Comparison of mean valuee for remains of
large and small animels ebendoned by predators al klllB
and Iransported by Ihem to denso 232
FigUl1l5.ZQ. Ccmparlsen of mean values for large antmals
found in predator dan, and found abandonad un
Nunsmiut Eskimo kili sltes. 233
Fl.ure 5.21. RlIlationship between lotal Nunamiul dll-
perled kllls and Ihe GUI fOl carlbou. 234
Figure 5.22. Relatlonship I:letween mean frequencies for
bon8ll of large prey tound in dens and the GUI tor
caribou. 235
Fl.ure 5.23. Comparison between plIr1stransporte<! for
,loraRe el Anaktuvuk vi!lege end mean values for parts
abv.ndoned on predator kills. 236
FIsure 5.24. Relationship ootwaen parls ramaining al the
"average" kili site for largeslzed prey and the GUI for
caribolJ. 236
Figure 8.01. Relationship between the percenlege of ex
pecled articuletor ends whlch are m!5sing end the count
of unldenl!fied fragrnents par MNI (or sites in Olduvai
Corse. 260
Fisure 11.02. Relalionshlp between the count o( unlden-
liflad fragmenls of MNI end !he numbar of cyllndafS par
MNI for sitas in Dlduvai Gorge. 260
Figure 8.03. Relelionship between the percentege of eller-
ticulator erlds represenled by complete bonas and the
number of cylinders per MNI for sites in Olduvei
Gorge. 261
Fig,,", 6.04. Ralationship belween Ihe nequencies af the
proximal and dislal humerus for sites in Olduval
GOl8e 262
Figure 8.05. Relationshlp between Ihe frequencles of prox-
imal tibia snd distal libia for si/es in Olduvai
Gorge. 262
Figure 6.06. Relationshlp between Ihe frequencies of
analomical parls fram FLK N, Level 5, alld modeled val
ues. 271
Fisure 6.07. Reialiunship between frequencies of analomi-
Lisl of Figures und Trrles
cal parts from FLK N. Levels 1 and 2. and modelad val-
ues. 271
Figure 6.08. Relationship between frequencies of aneterru-
cal parts from FLK N. Level 5. and Aleskan essemhlage
Irom Chandler. 271
Figure 6.09. Relationship between frequenctes of anatom-
cal parts Irom HWK E, Level 2. and modeled val-
ues 272
Figure 6.10. Relationship between frequannes of enatomi-
cal parts from DK. Leve] 2, and modeled vales. 274
Figure 6.1l. Relalionship betwssn frequencies of analomi-
cal perts from FLK. Level t S. and modeled vales. 275
Figure 6.12. Relationship between factor scores on
enetomtcal parts for Iactors 3 end 5. 277
Table 1.01. SecondllfY bomess end eslimatad prey ac-
cumulaled in vences habiteis 15
Table 3.01. Sequence of natural dismembermenl accordng
lo HilI 43
Tabla 3.02. Size d!strihulion of fragmenls recoverad from
ItIkmalaiyak wolf den 58
Tabla 4.01. Factor loadings for analysis of Nunamiut
faunal essernblages 93
T.ble 4.02. DistributicJIl of cut mllrks on bones in Uve
Nunamlul assemblagas 97
Table 4.03. Dlstribulion of butchering marks on bones
from Ihe Mousterien slle of Combe Grenal 99
Tab'" 4.04. Invenlory of described skinning and bUlcher-
ing marks 136
Tabla 4.05. Bone spllnlers recovered from six rnarrow-
cracking experimenls 164
Table 4.06. Frequency dlslribution of impact scara on ar-
ticulalor ends of long bones l6!>
Tabla 4.07. Vltriation in bone breakage for animal vllrsus
humanderived assemblllgllS 174
Table .5.01. Assemblage composition o( Alasklln wolf kills
(ceribou MNls) 211
Table 5.02. Andrew Hill's kili assemblage dala from East
Africa 214
Tablll .5.03. Richard Kleln's den data from South Af-
rica 216
Table 5.04. Survival percentages for bones of anlmals of
different ages 218
Table S.uS. Inventory for Il3semblages plolled in Figures
5.07 and 5.08 220
Table 5.06. RecolIslruclion of originltl Itssemblltge como
posilion for Alaskltn samples 225
Table 5.07. Reconslructed assemblages from
.swartkllp 22&
List of Ftgures und Tubies
Tabla 5.08. Summary dala (or human end nonhuman kills
and rmnaported hone assembleges 230
Table 8.01. Mary Leakey's classlficalion of Olduvai
sites 254
Table 6.02. Andllary facts: Olduvet faunel as-
semblages 258
Table 6.03. Olduvai Gorge data 254
Table 8.04. Behevoral rnodels for Olduvai as-
semblages 266
Table 8.U5. Sorted roteted factor loadings (patlern} for R
mode enalysts of Olduva sttes 257
Table 8.06. Factor acores for R mode analysis of Olduvai
siles 268
Table 6.87. Sorted rotated factor loadngs [pettern) Iur Q
mode analysis of Olduvai ettes 269
Tabla 6.08. Factor scorea for Q mode analysts of Olduve
sitas 270
Table 6.09. Comparison of destrucuon estlmales derived
Irom factor analysls and the humeros test. 279
T.ble 6.10. lnventones 01ertfecte recovered from factor 3
atea, Olduvai Gorge 281
*" 1
Note: Only the most commonly used cbbrevtctions are Jisted here.
ANT Antier
SK Skull
MAND Mandible
MAX Maxilla
AT Atlas
AX Axis
CERV Cervical vertebrae
mOR Thoracic vertehrae
LUM Lumbar vertebras
PF,LV Pelvis
SAC Sacrum
R Rihs
ST Sternum
se Scapula
H Humerus
PH Proximal humerus
DH Dislal humerus
Re Radio-cubitus
PRe Proximal radto-cubnus
DRe Distal radlo-rubltus
CARP Carpals
Me Metacarpal
PMe Proximal metecarpel
DMe Distal metacarpal
F Femur
PF Proximal femur
UF Distal fmur
T Tibia
PT Proximal tibia
DT Distal tibia
TAR Tarsals
AST Astragalus
CAL Calceneus
MT Metatarsal
PMT Proximal metatarsal
DMT Distal rnetatersal
PMAL 1 Ftrst phalange
PMAL 2 Second pbalange
PMAL 3 Thirrl phalunge
---------------------------_.- ..,..
The focus ofthis book Is bones. It
muy come as a surprise lo sorne thct
mosl o{the behaviorol ideas regcrd-
jng our ancienl post are dependent
011 the interpretotion of faunal re-
moins ond depositional context-c-
not. as mcst textbooks would leed
one to belteve, 510ne 1001s. In fod. in
aH the arguments 1 hove reviewed
regurding the behovior of ancienl
man, almos! 011 aredependent 011 in-
ferences drown rom faunal focts oc
focts of ossociotion between fauno
and stone 10015. Few jf cny plctures
of 'he post are, strictly specking. de-
rived from lnterpretutions o[ stone
A study of bones s, however, not
sufficient to c10rifyOUT oncient post.
We mus! kncw how lo use informa-
lion ond orguruae reseorch. and we
musl increcse the reJiability with
which we construct the post.
f will urgue thct crchceoiogists
hove regularly generured a variety
of modern myths by vrtue offaiJures
~ ~
Part 1
Documenting a
problem: Concepts
and observations
2 Por! 1. Documenting (1 Long--Ferm ArchoeoJogicol Problem: Concepts ond Ohservntions
in the inferential procese. Mnny such modern myths hove been genernted
by sertous and dedicated urchceologtsts through the interpretation of
founaJ mctericls. 1 am reienng to arguments regcrdlng the churccter of
life ct the dtm boundory of our humonity os known jrom the nrchceclogi-
col remoins of Lower and Middle Pletstccene homlnlds. There is a curren!
set of pictures o/ the post: the mighUy elephant hunters of Middle pJeis-
tocene Torralba, the cooperative altruistic sbormg behcvior af the
hominids al (he Pletstocene boundary. These unclent rnen are dcplctcd as
living in base ccmps. hunting, and in many ways resembling a bond of
modern Son speakers o/ the scutbern Africon veldt! Or if you prefer o
different view. the myth o/ the killer ape sees our earliest cncestors os
successful predators gaining Iheir livelihood from the flesh of other ani-
mals und possessing the "kler Instinct." Tllese modero myths abill!!
man ore bcsed on the interpretation of the o/ these
vlews may be uccurcte, bu! in the absence of a reliable methodclogy or
giving meaning to bones they must remcn myths based 00 judgments,
oplnlons. ond in some cases clear blcses, perhcps motlvuted by politicol
or humanislic phlosopbtes of rhe contemporary worJd.
II wiJJ become clear thct 1 do not hove new informatian or controJJed
data of direct relevcnce lo aH the types of interprefotive myths archaeoJo-
gists hove genercted. For instance, 1 wilJ only menon in pcsslug the
problems invoJved in Interpreng bone destructlon of human skeletcl
material by onimaJs ond other natural agencies, yet this is crucial te
evcluctlng the vchdity of argumenls for cannibaJism ond even to the
character of "religious" behcvior or the "spirituol life" of cncient mono
(See Beroumoux \1958) ond Marshack (1972) for examples of such nter-
prefive speculation.J Eveo fhe simple recognition o/ intentionaJ burial
and tne presence of graveside ritual is not on easy inference. For instonce,
(he NeanderthoJ "burial" of Teshik-Tash (Movius 1953) might just as
eosily be seeo as the consequence of destruction and sorting of skeletal
parts by predotor-scavengers instend of as ritual behavior by the Neaoder-
thalers. Nevertheless, the arguments to be presented regarding the fenu-
ous contextuaJ assumptions archaeologists frequenlly make, coupled with
the lack of controlled information obout bones, render many interpreta-
tions strongJy suspect. In short, their stalus as modero myths hecomes
more apparent.
Chapter 1 facuses on sorne of the assumptions that hove gujded many
of the reconstructive n(erences made from archaeoJogicoJ observations.
Chapter 2 focuses on the probJems of infereoce thot archaeologists must
squarely face i( they hope to increose the accuracy with which the ar-
choeoJogicaJ record is used as a meons to gaining knowJedge of the post.
lo essence, this port of the is concerned wifh two probJems: (a) the
understonding of the post tha! we wouJd Jike to improve Ihrough re-
seorch, and (b) the research strotegies that musl be employed if we hope
to achieve increased understanding oI the
As ts suggested in the Iille of this book. rny focus
is the study of fauna and the interpretations ar-
cheeologlsts offer regerdtng data summarized trom
observauons on fauna. Although tbe primaryinterest
s bones and the development of a methodology fm
using fauna te gan knowledge of the past. this work
gces beyond a mere technical concern for bones.
David Clerke 238) once charactenzed much of
rny carlier work as metaphysics:
Melaphysics .. the science Di being as such .. Ihe lerm
sdence is used in its dassic sense of knowJedge by
causes (Runes 1962: 19AI
Archaeological melaphysics is lhe sludy and evalualion
of the mosl 8eneral calegories i1ndconcepls wilh which
arch8eo!oRisls lhink IClllrke 1973: 121.
The eJtposure of archBeological melaphysics to critical
appraislIlallows liS the self-consciolls capllcily lo con-
sider Ihe possibilities of altering or rejecliog currenl
disciplinar)' concepls in favour of o.ll.,rnative forms
Thus. 0.1 th., mnment. archlleology is still a disciplinA in
which arlifllcls. assemblages. siles. and Iheir conleJtfs
lirA irlenlified Ilnd n,Jaled as relics of communilies in
a<;<;nrdance wilh rules formullllerl in terms of arlifacl
taJtnnomies--lhe Iradilionlll Monteliao parlldiRm
IClarke 1973 in Hammond el 01. 1979: \14-951
In Ihe foregoing sp.nse. Ihis book is directly con-
cemed wilh sorne asper:ls of archaeological
Chapter 1
Relics to artifacts
and monuments to

4 1. ReliGs lo cnd fa Assembloges: Chonging Fmmeworks

The Re/k and Monument Phcse


metaphyeics. I will advocate the replacement uf
sorne besc ccncepts, and modtcettone in the con-
venttonel use of others. In addtton. 1 will offer cer-
tetn methodologtcel suggestlons. which are tnevta-
bly celled for gven bestc shifts in metapbvscel
perspectiva. In fact, mos! oC Ihe book is concerned
with Ihe deveiopmenl of methods' however. the
need for such new methods s only mede olear in tha
r:ontext 01 changed perspectivas regarding
-chaeolosico.l record and the pas\. 1will be concerned
with the concepts crjcct. cssembicge. site, and
bcbtoncn. 1will attempt to demonetrete Iba! the as-
sumplions about the ercheeologtcel record that stand
behind the use of Ihese concepts are frequently mis-
leading; nevertheless, assumptions strongly condi-
tion what is considerad to be a problem in need of
solution as well as what is considered lo be a justifi-
eble slatement abou! the pest. More specifically, I
will be concerned wilh what the archaeological re-
cord is Iike and how iI comes lo be so. Given basic
conceplualizations about Ihe archaeological record
and in turn what the record means for the past, there
are linked melhods thel are generally taken for
The sludent of a discipline frequently reads and
uses Ihe products ofhis predecessors as discrela con-
tribulions from which he picks and chooses ideas
and observations lo serve his ends. This activily
should be carried out with an eppredation of Ihe
inlellectuel hislory of the field so the writings of ear
lier workers may be viewed against generalizslions
aboullhe thought of Iheir time. Sorne view of history
in the latter sense is needed to provide e pause before
the user of Ihe works of others begins selecting ideas
and observations for personal purposes snd/or ex-
pounding on Ibe value of past work relative lo pre-
sent points of view. A productive researcher needs a
wellfounded sel of ideas regarding the stele of the art
both past and present lo undentand why pest work-
ers chose to invest their time as hey did. He or she
neros to make judgments as to whol is in facl worlh
seeking through researeh effort al the present time. If
we gain sch a perspective on how Ihe state of the art
changes, Ihen the past deeds that we criticize need
not be seen so much as misdeed relative to our con-
temporary frame of reference but as imporlanl con
fributions necessary lo Ihe genesis of the point of
view froro which om critidsm derives.
There have been severa! histories (Daniel 1968;
WiIley end Sabloff 1980) of tbe atchaeologfcal disci-
pline. None have trnated ether Ihe fundamental or
mctaphysical propositions about the cbaracter of the
archaeologicaJ record, or tha sources or assumed
"causes" for the properties of the archaeologtcel re-
cord to which meaning was frequently attached by
the erchaeologtsts.
I wil! ettempt to charaderize uur Iield in funda-
mental Ierrns. that ts. in terms of the basie ideas with
which ercheeologtsts have worked. I will altempl to
iIlustrale how these ideas and Iheir use have con-
tributed to the current "state of Ihe ar!" while also
eonditioning the characler of many problems cur-
rently facing archaeologisls.
Viewed from a metaphvsical perspective, Ihe dis-
cipline has had Iwo phases, set apart Erom one
another in terros of the concepts commonly used for
treating the phenomena from which archaeologists
seek to learn of the past. 1terffi these Iwo periods Ihe
re/ic onrl monumenl phase and the ortifoCI and as-
semb/age phase. If sorne of Ihe current chaJlenges to
Ihe tenets of the arlifacl and assemblage phase prove
productive, a new phase may be in Ihe making.
The Relie and Monument Phllse

The reHc and monument phase of archaeological

invesligalion carne firsl and in many ways changed
our ideas of Ihe pasl mosl drastically. 11 pushed the
Crontiers of human history back from Ihe 4000-year
deplh of Ihe biblical perspective to Ihat of man as an
anden! resident o the planet. The intellectual pro-
cedures used in accomplishing this were based on
two propositions: (o) the definilion ofman as a man-
ufacturer of lools IInd (b) the definilion of lools man-
ufactured by roan as designed or manufaclured to a
plan-in shorl, exhibiting pallerned redundancy in
Given Ihese two propositions and the recognition
of the existence of slone tools (as opposed lo "thun-
der slones'" etc.), it was possible for a condusion lo
be reached regarding whetber or not lools had been
found, given Ihe evidence presenled by sllch men as
(183:l) and Boucher de Perthes (1849).
Demonstrating the antiquity oC man resled wilh the
strenglh of the assoctaton belween tools and extnct
animal remains wilhin undisturbed depostts. The
recogntton of the "hand of men." as it was frc-
quently referred to in the early literature, required
tbe demonstraton Ihat a design or plan had been
usad in the productlon ofthe tem (relc or construc-
tion Imonumenl} being investtgated. T'he place
where Ihe relie or monument was located was impor-
tant primarily for dernonstrang ossoctcuons with
natural obects. sueh as the bones of animals. 'I'he
lalter provided Ihe bass for chronolcgtcal assess-
rnont as well as sorne clue lo the nature of the envl-
ronments characlerisfic of the eras when maIl had
abandoned his ilems of handiwQrk. The invesliga-
lion of relies in these lerms ramained the primary
approaeh in Paleolithic researeh for a considerable
number of years. The investigalioIl of monuments
was largel)' aIl activity of c1assical archaeologists. or
atleasl those interested in Neolithic and more reeent
The early procedures were based on assumptions
characlerislic of the era in which archaeology began
lo emerge. Central among Ihese ideas was Ihal of Ihe
Creotor. A dislinclion was made between God's crea-
lions (nafure) and man's crealions lcivilizalions).
Tha plans of the "joinf' creators. Gad and man, were
lo be seen in patlerning manifest in their separate
crealions, nature for God. and relics and monuments
for mano
However. during {hese early years DE the relie and
monumanl era. !he role of God as divine Creator oC
nalure was severely queslioned. Wilh the debale
over Darwin's unprecedenled insight thal pallerning
in nature need no! be mferable to the prior execution
of a divine plan, but instead might result from Ihe
normal operalion of unplanned processes in nalure
itsel, Ihe equation belween God's design or plan-
ning and patterning was challenged and largely
abandoned. Vet man's role was nol quesfioned. Pat
lerning deriving from Ihe "hand of roan" conlinued
to be viewed as a malerial manifeslation of man's
plans and creative designs.
11 was in Ihis inteHeclnal context as applied lo
monuments Ihal important advances in archaeologi-
cal techniquF! were pioneered. Changas in method
during this period are generally attributed lo two
sourr:f>S: (o) cJa.qsieol archaeology (Daniel 196B: 65),
whieh emphasized Ihe detailed mapping of ilems
relative to ene enother. as well as in situ description
of what Taylor (1948) would later cel l "afnttes:" as
sources of nformaon about Ihe past: and (h]
monument-oriented reseorch initialed by Pitt-Rivers
at Cranborne Chese (published between 1887 and
189B). The tener served as a model for data-
collecng techntques. and formed the bass for a
conceptual shift toward Ihe investigalion of
"preces." a shift that would be soldted in the suc-
ceeding arttfect and assemhlage .pericd.
11 is not surprising that the development of tech-
nique was the product of monument-orlented re-
search. Excavations within monuments were inves-
tigalions within previously identified human con-
structions, such a hill forl. a long barrow, or a greal
palace. Within such complicaled constructions the
design or plan of Ihe builders could only be uncov-
ered by eareful excavation and recording ofthe many
parts making up Ihe overall conslruclion. The lech-
niques were not developed or employed lo demon-
strale tha! a monument exisled; that was assumed
The lechniques were viewed as proeedures permit-
ting or facilitating a more accurale description of the
manifest design or plan in terms of which the
monument was construcled.
Researchers of Ihe relic and monumenl phase
worked with several basic ideas and a melaphysical
that viewed the subjecl of study (lS things,
specifically relics (movable) and monumenls (nol
movable). These Ihings were recognizable as rele-
vanl lo man's past by virtue o Iheir demonstrable
organizalion of properties belraying a design or plan
that must have been in Ihe minds oftheir makers. For
much of Paleolilhic IIrchaeology. Ihese views were
coupled wilh lhe Victorian idea of progress, so Ihat
temporal Irends in increasing complexily, judg-
menls of refinement. or increased intelligenee were
taken as direetly reflecling lhe character of Ihe minds
or Ihe knowledge available lo those minds for de-
signing or creating Iheir products. The early re-
searchers inlo man's anliquity were concerned with
"man" in Ihe generic sense, and lended lo have a
single view of man as Ihe crealor of a progressive,
unilinear emergenee_ This view is well iJlustraled in
the following stalement:
Il was about Ihis period-lhe end of the 11181 cenlury-
Ihal a greallrulh-Ihe inception ofwhich we Dwe lo our
IJI1,;;,te... -.?-fv,A.(-J..,
-,--- A'
t. Relies lo Artifada and MClf1umenls lo Assemb/oges: Chonging Conceptual Frumewcrks
Thp I\rtifocl ond Assemblnge Phcse
cnlleagues of France-c-begen lo leaven the reseerches .
Ihis truth s simply the recognilion lha! the law of
'change al prosress, which tnuences al! the worldly
efetre of mano halda true no! only Ior the presenl bul
also for pes! Seneralions of mankind. Every generaon
has ita own dtsuncttve tesbtons and ideas: ji builds ita
hauses, it ttfls lts felds. it makes its implements , il
writes ita books. ir wears its clothes. and pants ts pie-
tures in 11rnanner alightly dfferent frum the generettcn
which went betore i11Keilh 1931: 466--489).
In addton lo generetton by generetton "pro-
gress," enalher idea gradualJy emerged duriog the
last part of Ihe nineteenth century: Ihe idea of plural
De Morlillel helped to show Ihal one tl8-
semblase of Iype-fossils oould be used lo define. no!
onlya pariodof archaeologicaJ limein france. bul alsoa
nlltlonknown lo wrilten hls"ny.. ,. It was however, in
NorlhernEuropeand especiallyin Germany . thal ar-
chaeologislS finl camelosea c1!:!arly Ihal assemblages of
Iype fossilsmight characlerizenol onlydislincl periods
of time, bul also dislincl nationsor Irlbeswilhina single
perlod. And il was German prehistorians who carne lo
terro such recurren! asemblages of lype fossils
"Kulturen"---eword... Iranslalecl nlo Englishas "cul-
lures" lChilde 1956: 281
These "discoveries" ser (he intellectual stage for
the developmenl of !he basic archaeological con-
cepts wilh which we generally continue lo labor to-
day. Cerlainly Ihey provided the basis fOla shifl from
Ihe search (or man's place in nalure to Ihe searc.h for
an appreciation of man 's ethnoeultural hislory: "The
mosl profitable inquiry {ofarchaeologyJ is Ihe search
for Ihe origin of epoch-making ideas in order lo
comprehend Ihe history of civilizalion 10. T. Mason
1893: 41131."
The Artffact and Assemblage Phase
A relie W8S a construeled objed, planned, de-
signed, and manufactured by mano An artifacl, on
the other hand, was any object or thing exhibiting
properties Ihal appeared to be Ihe results of human
adion. Relics and monuments were also things thal
exhibited propertes of Interest. Classifications and
texoncmes had reference ro varabilty in the prop-
entes manifes! amcng these individual units or
thlngs. With the conceptual shtft lo artifacls and as-
semblages a fundamental changa occurred in the
character of the baste archaeological units. Arlifacts
were recognizable by virtue of their cssoclnon wth
other objecre. in particular pleces. as well as by
vtrtue of a distincl design or evtdence of planning, as
was the case with a relk.. Stmilarly. assemblages
were only recogntzeble by vrtue of demonstreble
patterning in the ossccrcuons among things and
c1asses of things at ploces.
Although I have not atlempled to cover all the
lilerature of the period, it is my impression that Ihe
earliesl use o Ihe eoncept of arlifact (bul nol neces-
sarily Ihe wordl arose in two slighlly different re-
search contexls. The firsl contexl is Ihal of Old
World archaeologisls investigating relatively lale
remains where items included wilhin human con-
slructions (monumentsj could be viewed as being
there by virtue ofthe hand of man, even though Ihey
may not have been manufaclured lo a design or a
plan. Under such condilions an ilem or class of items
receives ils artifact slatus nol by virtue of formal
properties exhibiting "design or planning," bul in-
slead by virtue of its associalion wilh demonstrable
reHcs or monumenls exhibiting unambiguous plan-
ning and design within deposils considered lo hove
been construcled or at leasl formed "by the hand of
man," The second contexl in which the concept of
artifacl gained sorne importance was among
Paleolithic archaeologisls who frequenlly spoke of
recognizing prehistorie "habilations." When such
c1eims were made Ihe investigalors tended lo treal
Ihe contenls of "deposits" as if they were
monumenls construcled by Ihe hand of mano Sorne
of Ihe shift lo the metaphysics later associated with
the concepts of arlifacl and assemblage is seen in the
conlraversv between Ihe proponents of Ihe use of
fauna as eriteria for defining eras or periods in
man's past. and lheir opponenls, who advocated the
c1assification of relic.s and in turn the reeognition of
eras by virtue of the constellalion al reles systemali-
cally found in association at plar;es. By lhe lurn of
Ihe cenlury Ihis shifl lo recognizing assemblages was
generally adopled. (despite the fael tha! sorne par-
ticular relic classes were considered more diagnostic
than other fcrms]. In short. "bebtettons" were in-
creasingly Irealed as rnonuments. and the systemato
investigation of cave depostts ptoneered earlier was
pursued wth great tntereet.
It was in the Mousterien ege thet man first made hts
horne In caves: Ihe period takes ts name indeed from
Ihe cave of Le Mousfier in the velfey of Vezre, Dor-
dogne, where its remains were firsl carefully studted
[Sollas 1924:212].
Caves heve been repealedly frequented by groups of
m!:!n; during eaeh occupalion broken lmpl",menls, dis-
carded ornamenls. bones left over froro repasls. ashes
and olher refuse will be droppad and trodden inlo lhe
cavefloor.formingQ habilationlayeror culluraslralum
(freneh foyer), If Ihe cave be then daserled, ClIYe-earth
fal1s fromIhe roofor slala;rnlle mayforma sterile[ayer.
The laller will seal the underlying habilalionlayer. bul
roay forma floor on whlch man, eYenlulIlly relurning.
will deposil a second occupalion layer.. Plalnly lha
artifacls and other dala in aaeh deposit conslitute en
tlssemblage of assocaledlypes IChilde1956:59J.
The shift lo Ihe treatmenl of a habi1ation as if it were
a monumenl braughl an interesting change in !he
characler of the criteria used to identify the habita-
lion. In the earlier periad one recagnized a reHe
and/ar a monumen! by virlue of the plan or design
manifest in Ihe organizalion of the thing itself. Habi-
lalions weTe not so easily recognized, particularly in
Paleolithic research. HabitaHons were instead denti-
fiable by a characteristic pattern of association
among lhings al a particular place. Thal is, recognt-
tion of habilations was in terms of a diognostic 05-
sociolion among Ihings al particular places. This
provided Ihe basis for a shift nol only in Ihe unils of
analysis and synlhess. but also in Ihe fundamental
character of Ihe units of observation. Whal were lhe
crileria fOl reeognizing a habitation?
They were as folJows:
1. The aggregaled presence of demonstrable re-
Hes in the design sense of Ihe term
2. The associaled presence of animal bones as-
sumed to he indir:ative o( Ihe meals of andent
J. The associated presence of fire
4. The assor:iafion of the foregoing eiements in a
"sheltered" setling. as inside a cave or
These eharaeteristies provided an operalional defini-
tion of a habitefion Ihal was drawn from our con-
tempcrary human experence of living in hornes or
resdences. It W<lS consdered natural for rnan lo
bring food IDhis hnma, lo Increese the comforl rector
by warming the place with re and by eilher living
in sheltered locettons or constructtog sheltee in
naturallv unsbeltered locatlons. If the relic and
monument period worked from a mlnma! definition
of mcn the tcolmuker; the erttfect and assemblage
period worked with en addtional definilion of man
the homemaker.
It was in the inle1lectual conlexl of bolh of Ihese
ideas of man fhat the prabJem arase thal is addressed
by ihis book. Nomely, how do we recognizeJ!!f!
of human --hominid beha.... ior in fa_upol re-
moins? This problem arose early in the history of
archaeological research and was addressed boih
from the perspective of attempting to identify reHes
and from Ihe slandpoinl of viewing remains of
human activities conducled in man's "home"-thal
s, from an associalional perspective utilizing Iha
concepl of habitation.
The two slrategies used and argued over have
been slraightforward. According lo Ihe (irsl, Ihe
morphological properties of poslmortem modifica-
lions on bones show evidence of design and plan-
ning and lhe modifications are therefOle referable
lo the hand of mano This strategy is a paltern
recognition approaeh, in which one firsl altempts to
establish Ihe exislence of a paltern 01modification,
and Ihen argues thal the form of patterning isolaled
is referable only to Ihe actions of mano This is the
same strategy Ihat has bellO widely researched and
applied to the diagnosis of rnodificalions on slones
and in tum Ihe recogniUon of stone lools.
Earlv in the reHc and monument era, claims were
made bone lools based on !he recognilion ol pat-
lemed modifieations and the referral of sueh pat-
tems to human agents. SimiJarly Ihere had been the
recognition Ihat animals modified bones and thal
many modifications Ihal were uncritically attributed
lo Ihe aetions of man eould we1l be referabJe to the
Bctions of animals. 80th arguments have been pre-
sen! from Ihe very beginnings of modern archae-
olagy. For instance, Boucher de Perlhes (1849: 312J
reporled B phalange wilh a round hole in the shaft,
whieh was discussed as 11"whislle." Oean Buckland,

8 1, Heftcs to Artfocls ond Mnnuments lo AsspmtJlup;es: Changing Conceptual Frnmeworks
The Artifuct end Assemblage Pbcse
as eaely as 1823, reportad on the similarily between
bone modification observed in depoaits beheved lo
have been accumuleted by hyenes and abone exped-
mentally gtven lo a spotted hyena at a traveling
menegerie. 80yd Dawkins. writing during the forma-
live years ofthe relic and rnonument era, commented
un Bucklend's earlim work as worthy ol emulation:
Dr. Buc:k1and's method of solving the problem o 'he
nnoducuon of rema los of so meny and differenl ent.
mals tnto so smell 8 spece. a a model of scientlfic
analysis. He arguea (rom Ihe abundance of 'he remains
of Ihe hyaenllllTld (roIDthe correspondence of their lee!h
with Ihe marks on ,he bones, lIod from 'he quantity of
Ihair coprolites. Ihal the cave was inhabiled by mallY
gBnerationsof those animals, and Ihal the gnawed frag-
ments were relks of their prey.. '. He further demon-
slrated Ihe truth of his conclusions 1' the crudal exper-
ment of subjecting the leg.bone of an ox to a spotled
hyaena from the Cape of Good Hope, in Wombwell's
menllgerie. the exact correspondence of one of he
fragments of the libia of an ox gnawed by lhe Cape
hyaena, with Ihe corresponding bone of lhe bison from
Kirkdale,. . The same kind of identity runs through
the whole series of bones gnawed by the living and fos-
sil hyaenlls.
Dr. Buddllnd's conclusion that the Kirkdale cave was
he den of Ihe spolled hyaena (H. crouto) lhal preyed
upon the animals ofYorkshire in ancienl limes, md Ihat
It was undisturbed down to the lime of ils exploration.
canllot be disputed. The trelld of the hyoenas in their
passage lo ond fro had polished lIome of Ihe bones and
jawsscattered on Ihe floor IDswkins 1874:281-2841
The recognition Ihat animals and olher nonhu
man agenls rould modify bones to a pattem promp-
ted sorne early and importanl research designed lo
provide recognition criterio for modificalions af-
fected by humans as opposed to other agenls. This
was considered Important. since c1aims for "Tertiary
man" had been frequently made during the lasl half
of the nineteenth cenlury, based nol infrequently on
claims lhal bones Ihal had been modified lo a "plan"
had been found in early deposils. In fac!. approxi-
mately the firsl 50 pages of De Morlillet's famous
work Le Prhistrique (1900) are concerned exclu-
sive!y with the recognition of bones morlified hy
man as opposed to animals and other
What I find e)(lremely inleresting is lhe second
researeh stralfJgy followed in idenlifying bones mod-
ified or inlroduced to deposits by man opposed lo
other agents. The most eomman strategv was to
sludy the patternng considercd manifest 00 bones
reccvered from hebttattons. Snce man was res pan-
stble for the depost he rnust aleo have been raspen-
sible for Ihe pattemlng manifest in and on tts con-
One of the earhest systematic descnptions of
modifed bones recovered in an archaeologtcal set-
ting was Ihat by Henri Martin (1907 -1910). His work
is a mixture of salid observauon and descnptlcn.
which are sUB useful. and a kind ot speculation that
seems Hule more Ihan romanlic faney. Part of the
"faney" derives direclly from Ihe assumplions and
reasoning Ihal underlay Ihe pallem-recognition
studies he conducted. Beeause 1he material he was
anal)'z;ing carne from Ihe archaeological sita of La
Quina. which was assumed to have been a habita-
lion, the contents of Ihe site were accepted as being
Ihere by the hand of mano Consequently, patterned
modificatioos were thought lo be referable lo human
behavior. In short, one assumpfion about the charae-
ter af the archaeo10gical record-that Ihe site was a
habitalion-was being used as the control condilion
for referring Ihe patterning observed in Ihe conlenls
of Ihe deposi1s lo human behavior. This approach is
well ilIustrated in one artide by Martin {1906) in
which he suggested Ihal a reindeer melapodial Ihal
bore the unmislakable impressions of animal teeth
demonslrated Ihal Ihe occupanls of Ihe site of La
Quina musl have domeslicaled Ihe dog, for in whal
other way could Ihe marks made by a beasl occur al a
human habitation?
Many of Ihe early inlerprelalions of andenl IIfe
were thl'! consequences of a similar logic. A good
example is provided by the Sirgenstein cave. exca-
valed just after the turo of the cenlury (Schmidt
1912). of particular interesl was Layer C. which
yielded a Mousterian assemblage associaled with
large numbers of cave bear bones. This mixing of
lools and bones was ciled as evidence thal Ihe cave
bear bones did not occur in Ihe context of a den
deposit bul in facl represented bears Ihat had baen
exploited by man as food (Schrnidl 1912: 165). In
addition lo Ihe large quantities of bear bones. there
were al leasl two "rodent layers" lhat yielded no
cultural materials. On top of Ihe Mousterian mate-
rials was a layer yielding Upper Paleolilhir: 10015 as-
sociated wilh a flluna dominafed numerically by
hyena. Despile whill appears lo me lo 00 provocative
evidence of a palimpsest depostt, deriving from a
variety of events and acttons by both man and ant-
mels. the excavators tnterpreted tha enttre deposit as
the result of homntd behavior. Man selectvely
hunted bears end later hyenas!
This type of approach was eommon end was not
infrequently associated with the clams for the use of
bone as a tool. The irnportant cave of vogelhenj Lit
Wrtemberg, Germany, is an example. The site was
excavated in 1931 by Oustav Rlek, who published a
three-volume repcrt [Riek 1934), The cave had not
been dislurbed by modern excavations prior to in-
vestigation by the archaeologisls. Nurnerous layers
were recognized. bul mosl of the bone malerial as-
sociated wilh Mousterian tools was recovered from
an area described as oval in shape just lo the left and
behind the southwest entrance. The largest bones
were said to be arranged along the cave walls; the
smal1er bones. in Ihe cantero The Mouslerian lools
were described as lorgely restricted lo an area 3.5 x
4.0 m, where the artifacts and bones occurred in
"c1ose contact." Cave bear, cave hyena. wolf. and
cave Hon bones were Ihe most common in this "close
association." where the approximately 50 stone tools
were recovered. In IIddition to the stone tools. a
number oC modified bones designated 10015 were
recognized by the excavalors. These were Iisled as
seven "bellscrapers," three "hip-joinl-sockel scrap-
ers,,' Iwo "smoothing" lools. and one awl.
Bellscrapers have been described as "Ihe libias oi
large animals such as Rhinoceros, horse or '.<lUle.
The tibia was broken in half. . [andllhe sponge-like
insides of the bon!!s were more or 1ess complele!y
removed so as lo produce a hollow space [Zolz
1951:931." This is exactly wha! has been described
by Kitching as "tankards orscoops" (1963: Plale 25)
and flaying tools (1963: Plate 30). Andree 11931:372)
reports Ihal Ihese same types of ilems were found in
Ihe Lindenlaler hyena den and olher German siles.
The "hip-joinl-sockel scrapers" are described
(Zotz 1951:93) as "joinl sockels of the cave bear. The
apophyses of these sockefs were partly removed, and
the sockel edge shows work Iraees in form of
moolhness alld splinlering."
A more expanded list of "bone lools" from Ihe
Renis cave was givf1n by Hulle (see Andree
1931:406), Numerous lt!vels were recognized and the
Jower Iwo "cultural" layers, Renis 1 and Renis 11,
yieldcd Mouslp,riall hlallspilzen materials in assoda-
tinn wilh a fauna domneted by cave beer in Renis I
ami both cave bear and cave hyena in Renis Il, Oc-
currng in both levels but reported lo be more com-
mon in Rens II were a number of "worked bone
tools." In addition lo the already descrtbed bell-
scrapers and hip-joinl scrapers were so-callad "fur-
smocthers.' and "worked" cave beer mandbles. The
identity of IheSE! items as tools has been repeatedly
questtoned (WelzeI1969; Zotz 1951:136), neverthe-
lees. the descripton ofsuch "10015"Irom levels dom-
nated by hyena and cave beer remains s interestmg.
Sorne of the more irnaginative and fanciful "bone
laols" have been described from famous bear cave
deposits. Particularly important in this regard is the
report on Petershole by K. Hormann (1923). A total of
20 stone tools are reported from three separate depo-
sitional levels in Ihe cave. Large numbers of cave
bear bones occurred throughoul the deposits in vari-
ous side chambers. These were described as piled in
crevices, on stone slabs. and in interesting combina_
tions. In addition lo lhe cava OOar Ihere were 55 iden-
lified elements of carnivores from Level l. Cervids,
bovids, and horses were represented by only48 iden-
tified elemenls (see Htirmann 1923: 137-138J. Staled
anolher way the assemblage was dominated by Car-
nivora. There were few stone 10015. bul a relatively
large number of "worked bones" were reported. One
of the more imaginalive forms, "bultons." are ilIus
traled as small see/ions of long-bolle cylinder. In
addition. Ihere were a number of "bolle knives," de-
scribed as sectians of bone with chipped retouch
along ends or laleral margins; "spearheacls and ar-
rowheads." identifiable by their suggestive shapes
only; and finally a number of "worked bones" ex
hibiting radial scarring, lateral 'relouch." and smoo-
Ihing of ends-all of which are indistinguishable
from fragments of humeros identified by Kilching as
"knives. pounders and scoops." {See Plate 17 in
Kitehing [19631 and compare Plate XLVII in Hr-
mann 11923l.} Many of these sorne "Iypes" of
"worked bone" have been reporte<:! from more re-
cently investigaled localions such as Repolust-
Hohle (MotU 1951:23,33-34. Plale XXIII), where the
fauna is likewise dominated by Carnivora.
Although the Cerman and other North European
Iiteralure cerlainly has many I:1xarnples of wha! cal]
be called anlhropocentrisTTJ-lhe assumplion
thal if any evidence exists for Ihe presence of
homnids everylhing present is Ihe product of


10 1. Ralles to Arlifor:ls und Monuments lo Assembloges: Chonging Conceptual Frnrnewcrks
The Arlifoct ond AssemblogePhcse
homlnid ecton-c-some very explicit examples come
from Great Britain. For tnstence. the famoue Kent's
in Devonshtre, England. excavated by the
Reverend Iohn MacEnery end leter by WiIliam
Pengelly, exhibits all the properttes ot having beco
primarily a den of cave bears sud then later a hyena
lalr [sea tooth counte in Cempbell and Sampson
[1971; 15). Mixed with the faunal remetas were
Acheulian sud Mousterian tools. The epetel [ux-
_--t-aposition wes sufcent lo leed Campbell and
Sampson {1971: vii) lo suggeat tbat "specieleed
butchertng al htbemaung bears durng wtnter may
haya been practil..""t!d by the so-called 'unspBCalized'
A<:heulian hunfers."
Tha dominant Une 01reasoning used in defining
bolh patterns "diagnostic" of human behavior and
assodations ol ltems thal served as the basis for be-
havioral nferenees was as follows:
1. The argumenl was made that an archaeologi-
cal sile was a habitation.
2. A habitation was considered a type of monu
3. The assodaled elemenls within a monument
were all Ihere by virtue of the hand of mano
Q. Any patterning demonstrable within
classes of assodated elements was reler-
able to human actions.
b. Any patterning demonstrabJe among
c1asses ol 85Sodated elements was refer
able to human acUons.
Tftis parlldigm of reasoning permitted sorne "recon-
struction" of the past based on observalions made on
the archaeological record, and as a result of its use
several ideas about the past were gradually sol-
idified. There was, for example, periodic f1irlalion
with the idea that man may have made and used
bone tools befare he used stone (see De Mortillet
Menghin 1931). More important lo the
futura directions of Paleolithic archaeology were
three recurrenl arguments about man's nast:
1. Bone tools were common ond fraquenl com-
ponents of MiddJe ond, perhups, Lower Poleo/ithic
1001 assembloges. They were recognizable by their
shapes, which were suggestive of their uses, liS well
as by superficial modificalions and forms oC break-
age. Advor.Jles of Ihis view were Henri Martin
Sehmidt P912), Riek 11934), Andree
(1939), Hrmann (1923), and Bachler (1921, 1940).
2. There were dfstmcttve bone too/s cssccrcted
with the remotns of cave becrs as weJ1 as evidence
jnr the special treotment of cave becrs und theh re-
mcrns. This vlew ergued Ior the extstence of a bear
cult and proposttons were advanced regerdtng the
spntuel characterstcs of Middle Paleoltthtc mano
These argumente may have developed Irom views
popularized through the work of Schmidl and Wer-
nert (1910). Schmtdt'e work has been summanzed by
Sollas (1924), who wrote,
Of htte years Germany has added i1s contnbulions, and
one of the mosl precisely investigated caves of modern
times is that of Sirgensteln . the broken boMS of ani-
mals ware strewn about. The Cave Besr (Ursus
speJoeus).. was byfar the eommonest; this animal was
tha universal "caretaker" oecupyingthe cave during the
absenee of tha hunters snd reeeiving shorl shrift when
they relumad.Jt was alsoa favoritefood, as is shown by
the great numbers of bears' bones whieh are massed
together near Ihe threshold of the cave as well as plenli-
fuJlyscattered about. Noone cave could have supplied
so manybesrs. and the hunters musl have ransackedthe
surrounding dislriet in pursuit of Ihem. Mosl of Ihe
bones belonged toyoung animals whieh were. no doubt,
an easler prey as well as mole deBcateealing [pp. 221-
In these early works there is Ibe nolion of an al-
most spedalized hunting of cave bears. In lhe mid-
1921ls this emphasis or suggestion of a special rela
tionship belween man and cave bear was elaboraled
by Bachler (1921,1940) and Hrmann (1923), and the
idee of a bear cul! emerged (see Abel 1925). In the
words of later wrilers presen\ing "general under-
standings" of the pas!.
Why Neandarlhal man began "unting tha Cave Bear s
not cerlain. It was 11formidable animal. slanding mOfe
thllneighl feat laJl when reared in anger. IInd musl have
been a dangerous foe. It also Jived in mueh more lnae-
.::essible pl!fces than mosl of the Olhergame, Neverthe-
les'. iI WllS hunted-perhaps lo fulfillan early hunting
rllual Discoveryof bear skul1sslacked in a slone ehest
in Orachenloch. Swilzerland, supporls lhis idea; Ih.,.
heads mayhave been Itophies IHowel! 1965:1261.
In high-I.,.vel eavllS of Swilzerlillld,Bavariaami Croatia,
evidenee exis!s of a cull of beara involvin: the deposi-
lion of besr skulls and bones in certain raves .. In a
chamber of the Drachenlochin Swtzerlend. a stone ctst
had been built lo hnuse stecked bear-skulls; piles of
scrted long borres were laid along the walls of fhe cave.
Another heep of bcnee contained Ihe skull of a beer
througb whch a lag bone had been Jorcad, the skulJ
resting upon two other long bones, eech bone from
a diffelent beest [Colesand Hlgga 1969:286-2671.
This idea of a spectel celatonship between man and
cave beer seems to have struck a responsive note in
modem readers. for almos! all Ilterature treettng the
"origine of religin" cites these argumente and
pleces particular emphaes on thesa "important" de-
velopments leading lo our "human" status.
3, Ancient man wos [} cannibaJ. Perhaps one of
the more inOuential early c1aims for ancient can-
nibalism was made in the repar! on exca:vations at
Krapina (see Gorjanovic.Kramberger 1905). Other
aulhors during the same era sounded similar notes
[set:! for instance Rutot 1907). These early sugges-
Iions were greatly elaborated both prior to and im-
medialely after World War n. The cannibalism
suggestion was gradually linked with lhe so-called
"Cult of Skulls" (Bergounioux 1958; Wernert 1948).
For instan.::e, Weiclenreich (1928) describes the
Ehringsdorf skull in the following lerms:
Ihe frontal area beara undoubted signs of havng been
hil by stone weapons and it can be assumed that Ihis
individual had heen murdered. (b) Ihe absenee of Ihe
eranial base. no fragmenl of which was conlalned in Ihe
block of travertine fromwhich Ihe skull has baen 6)(-
Iracted, indicates Ihat the skull had been opened to ex-
traet the brain lp. 671
Laler Berckhemer (1934) described the Steinheim
skull as having been artificially widened at the base
around the foramen "for purposes of extracting Ihe
brain." This view was furlher supported by Ihe de-
scription of the Monte Circeo skull by Blanc (1939).
This important skull was recovered from a small
chamber nked to several other cave rooms by nar-
row passages. In all, the complex can be viewed as
seven caves. five of which are reported lo have
yieJded some MOllsterian remains. Blanc has re-
portad on this find in numerolls places [1950, 1958,
1961]; the skllll is said lo have been Iying as an iso-
Jaled element surrounded by rocks near the center of
a small chamber. Aones of olher animals are reported
concenlrated in al least Ihree addiliOlllJ locations
within the chember (see Blanc 119581 for a good de-
The lhree basic ideas just described largely grew
from early twentieth-century exceveona of Europeen
"cave" and rockshelter deposita referable to men of
the Mousterian era and of more recent times. The
exctement of a first glmpse of men considerably
earer than the Neenderthelers carne with the dis-
covery of the slte of Choukouten in China. This site
was remarkable for Its greet antiqutty, and be pree-
ence of importan' fossil hominid rematas. It was also
extraordinary in its being a cave deposit that yielcled
tools, evidence of fire, and bones in vast quantities;
in short, it was viewed as a habitation site in the
sense in which the European caves had becn viewed,
The methods of interpretation already common in
Europe were used to make slalements about man's
behavior in the remate past as well as to provide the
"control" Cor recognizing patterned produets of
man's aneienl handiwork.
In 1935 in aslo the Abb Breuil delivered 8.11 in-
f1uential paper in which he propasad that there had
been an extensive bone tool industry al the ancient
Chinese site:
Lel's go hack in Ihought, than. fa that dark dawn of the
industrial phase of mano We find him or his intelligent
ancestor surrounded by animalabatter armed by Nature
than himself, Iions and bears, with teeth and d8WS.
grass-ellters wilh homs and What more natural
lhan to rob fhem of Ihese wellpons to use 8gainst lhem?
Ever a hunter. Manhlld around him the ,keletal remains
of his vctims, quickly unfleshed by himself, by carni-
vares and by natural agencies.
He also encounlared on his wanderings Ihe east antlers
of deer and Ihe carusses of lhe carnivores' prey. Sorne
of !ne completed longer bonas made Itxeellenl clubs
witn handlllS not ee.sily broken:some of Ibe boneseould
in thair natural slale be used for
holes; some larger and flal eould be used as shovels;
others which were lonller, ltS levera; yel others whch
were shorl and sloul mlght suggest an IInvil or a h8m-
mero while such as were broad, thin snd trenchan\,
when grasped. became scrapers and
When Man. armed with 11 pebble. broke long bones to
exlracl the marrow. like a hYllena, they broke in dif-
fprenl WII}'S. Sorne had arlkuhtr ends eonvenient for
holding if man wishen to them as lmplements,
while the olher enn was pointed. Others. pieces of

1. Hehcs lo Arlifacfs and Monuments to AssembJalles: Changing Conceptual Fromeworks
The Arlifoct cnd A$sembloge PhaSe 13
diephysea, had ends that could be used as points or
chisel, end culting edges which could serve as knives or
scnrpers. Vnlike tha hyaena Man often broke the borres
Iengthwise lBr8uiJ1936:57-56; emphaais added).
Here we see the imaginativa picture of bone-uslng
hormnids that later captured the fancy ofmany audio
ences as the "osteodontokeratic culture" of yet aar-
lier man advocated by Rayrnond Dart during the
1-950s.However, tha history of ideas dealng with the
Choukoutien material is in itselffascinating, beceuse
ir tllustretes opposlng interpretive ponte of view re-
garding bone too1s, views slill with us today in the
old World and increasingly becoming an active
issue in New World studias (see Friedman 1980:7).
Breuil {1939) studied the animal bones from
Choukoutien, and Weidenreich (1941) reported the
hominid rernains. In the works of Ihese two men,
parallal theses were presented: (o) that an exlensive
bone tool industry existed al Choukoutien and (b)
Ihal "Sinonthropus," Ihe "man" responsible for bolh
stone and bone industry, was a cannibal. The argu-
menl offered by Breuil was based on Ihe suitability of
Ihe shape of abone for use in a given manner, with
Ihe inference lhat it would, Iherefore, have been so
used, and the modification of bone as evidence Ihal
intentional shaping by man was involved. Breuil
ciled the presence of rounded and "polished" edges.
chipped or tlaked edges, incisions or abrasions on
flal surf8ces, and "characteristic" longitudinally
broken bone. Weidenreich's argument for the can-
nibalistic consumption of bone marrow was based
on fractures of Sinonthropus long bones. "Long
splinters del off lhe shaft longitudinally so Ihal the
medullary canal lies exposed" were considered
characteristic of the actions of Sinanthropus: "This
manner of splitting is generally believed lo be the
result of human manipulation carried oul with the
aid of implements in order lo exlract the marrow
IWeidenreich 1941:751."
In the work of bolh Breuil (1939) and Weiden-
reich (1941] tnere was Ihe clear belief that sorne
long-bone breakage pattems were distinclive and
even diagnostic ofhuman agency. In addition, Breuil
interpreted modified bones as resulting (rom human
In Ihis era fhere was, however, rclevant research
that provided evidence lo the contrary_ Weidenreich
cites a study by Zapfe (1939:761. who experimentally
studted the way in whch living camvorea, espe-
cially the hyena. cbew long bones, and even ilIus-
treted how sorne of the broken Sinanlhropus long
bones corresponded lo the propertes cted by Zapfe
as charectertsttc of carnvores. Howevee. he con-
cluded thal "Ihe possibility extsts that the long
femur splinlers were not the produet of Smon-
thropus hirnself. This alternartve. however. seems to
me very unlikely Ip. 771." In other words. despite the
evtdence. wetdenretch preferred to vtew Sinan-
thropus as a cannibel!
Nearly the same sttuaton exista with the argu-
ments presented by BreujJ. Wen-chung Pei (1938:1)
then director of the Cenozoic Research Laboratory in
China, published a very important monograph,
largely in answer to Breuil, rightful1y posing the
question. "Commenl reconnaltre I-action bumaine
sur un debris osseus?" Pei proceeded to present the
information available to him as lo the modifications
produced by rorlenls and predators, as well as the
effects of water, soil chemistry, and sol lexlure or
bones. Pei was convinced Ihat many of Ihe
modficalions cited by Breuil as evidence for human
use appeared ret"erable to other agents.
After World War TI [1939-1945) many of Ihe pre-
war argumenls were continued (see Battaglia 1953;
Ehrenberg 1959, 1962) regarding the existence of the
hear cuH and associated dislinclive "Alpine
Mouslerian" (Leonardi 1958). The focus of con-
troversy regarding bone tools was. however, laken
out of its regional and provincial contexl with the
daims by Raymond Dart (1949) for an osleodon-
tokeralic culture anceslrallo the production of slone
tools by our earliest hominid anceslors. Dart's argu-
menls have been reviewed by Wolberg (19701. These
c1aims placed the conlroversy over interpretalion of
modified bones al the very of human hislory
and drew Ihe atlention of all regional specialisls as
well as a widespread lay audience. Dart's claims for
bone tools in lhe australopilhecine sites of South Af-
rica provided paleoanthropology wilh a Ii ...ely and
cerlainly imaginative controversy, which has con-
tinued to the present lime.
Darl's argumenl was very similar to that of Breuil.
He ciled the polenlial utiJ ily of Ihe shapes 01' various
anatomical parts and Ihen inferred Ihal early man
must ha... c taken advantage 01' Ihis potential. Adual
modcatons cited as diagnostic of human acuons
includsd an allegedIy distinctive mode of long-bone
breakege. the "crack-end-twtst'' method resulting in
preshaped bones and spirally fraclured bone sptn-
ters, and (as in the case of Breul] varous abrasions,
strtattons. and nctsons. as well as chipped or
"worked" edges. In addition, Dart made a slrghtly
different argurnent. namely, Ihal evdence for rnan's
selective use of anatomical perts rested wth the di-
ferential frequencies of "useful" parts in the sttes of
early mano Thus. for Dart two important seis of Iacts
converged: modlcauons on bones. which were in-
terpreted as produced by man. and the assocated
differential frequendes of anatomica) parts, which
were interpreted as resulting from the differential
transport of hunted foods lo "man's home." For Dart
there were facls of modificolion and foclS oJ as-
sembloge composition, which he believed could be
dled as evidence fOf hominid behavior. It is not sur-
prising that. given Dart's claims, olhars would offer
as evidence modified andJor broken bones in supporl
of the position that 1001-usiog hominids occupied
different places other than Africa prior to the de-
velopmenl of stone tools. Such c1aims have been
made for the paleontological remains from the
Netherlands (Luttschwager and von Remmel 1962)
as well as Rumania (Nicolaescu-Plopsor and
Nicolaescu-Plopsor 1963), Many of Ihese arguments
have been seriously challenged (see Feustel 1969,
1970); neverlheless, they persist on the level of
"opinon." Thal is, specific modificalions are com-
pared. and then the opinion is offered c1early
only aman could have produced the modifications
With Dart's c1aims we Ide an importanl step away
from Ihe propositions that permitted the recognilion
of a habilation during the firsl half of the twenlieth
century. Mosl important was Ihe absence of stone
tools as a criterion or concomitan! in his sites. He
sought lo meel Ihe tool criteria by arguing for Ihe
presence of bone tools. Dart, obviously knowledge-
able about the earlier work that described
morphological patterning among bones assodated
with unembiguous slone tools al habilalions, could
now appeal to thal patterning to support his daims
for the modified and broken bones from his sites as
tools. This c1aim lent support to his fmther daim
that the sites were hahitalions. thereby justifying in-
ferences from tncluded malerials about the behavior
of AustmlopHhecus. Dart's case ls of no small sg-
nicance. since his early claims for the ancestor
status of the Taung skull to modern man had largely
been rejecled on morphological grounds. Dart in-
creesingly argued that the eustralopithectnes were
protohuman by virtue of behavioroJ cherectenetrcs
manfest in the archaeoJogical record. This was the
firsl Importen demand or behovioral informal ion
from the orchoeoJogicoJ record of unclent mano Prior
to Dert's claims there had been some altempls lo
charactenze Iha behavtor of early man. but al no time
were the inferred behavioral characleristics consid-
ered eilher crucial lo argumenls abaut Ihe sequence
of human evolution or central lo the important ques-
tion of the hislory of "human nature," as was Ihe
case for Ihe behavioral daims by Dart 1949, and
especially 19571:
Ihe innomlnale bone found al Slerkfonleinand Ihe oc-
ciput fromMakapansgllt hovepruvldedfurlher evidence
lhat these creatures artlmorec10sely related lo man Ihall
lhey are to the ape$ They are prolo-human belng5and
Iheyare the most primltlve hominids of whom we huve
knowledge. JI therefore becomes ur-
genl that we should elicil fmm those in which
prolo-humans occur alllhe ovoiloble information beor
inR on their manual dexlt'rily and implementa/ inlelJi
gence IOarl 1949:2:emphasis added].
The questions being put to archaeological record
of early man in Africa were not hislorieal in Ihe
sense of !hose asked by archaeologists working with
remeins from more recent periods. They were con-
cerned with fundamental questions about Ihe very
nature of "protohuman" behavior; Ihat is, how do we
get behavioral information froro the archaeoJogical
record? In Ihe conlext of this importanl shift of em-
phasis and in tum its demands for information about
the pas!. sorne important changes began to take place
in Paleolithic research.
ln the era immediately after World Ww 11 a sel of
c1aims was made by Africanists, which foreshadowed
a long and importanl sel of arguments relating to
Ihe nature of earlv roan's behavior. rhese were the
c1aims by Louis Mary I.eakey that Ihey had iso-
lated living floors at the site of Olorgesailie. Both
l. Desmond Clark and Merriek Posnansky have ex-
pressed Ihe opinion to me thal the search for living
,. l. Betcs lo Al1ifocr:s ond Monurnents to Assembieges: Chonging ConCl.!ptuol Fmmeworks

The Artifoct cnd Assembloge Phase 15
TABLE 1.01
Secondory Diomo" ondEsUmoted Prey Accumulatedlar
Vorlous Habltat:;;"
Data forllorthetn habitat.from McCullllUgh 1910; datafurtemp-
ateand tropical habita'. from bourllen 1963.
-u.ta ref..r lo carlbou Ofly
'Oata retarlo Ullflulatlll only.
Oata refl:tr lo aOO primatea unly.
150,000 to 200,000 anmals tktlledl ennuelly." In
other places predators remove an estimated 9-10%
of the total ptey btomass ennuellytscheller 1972:391).
As a simple axercise. let us see what such a situation
would look like if we essumed that predetcrs are
equally successful in all habtats. Table 1.01 presenta
summary dala on the emount of secondary biomass
that might be expected in different envtronments.
aesuming equal success of predalors [not very
likely-we can expect that as food density goes
down predator effectiveness may be reduced some-
what: certainly there would be a dependence on
smaller animals], AIso the amoullt of pr"y (dead
animals in kilograms} that might be expected in the
different habitats per square kilometer.
What is clear is tha! we may expect maior annual
accumulations of bones in tropical savannas, sub-
stanlial accumulatlons in temperate grasslands, and
minar accumulations in temperate forests, tropi.
cal rain foresls, boreal lorests, and tundra. Just as a
demonslration, lel us imagine 8n environment aIong
the grsss margins oC s boreal forest such as one might
find in central Manitoba, Ganada, Dr in numerous
mountaln s"ttingll. Let us propase that predator kllls
would be roughly the mean of expected values for the
. ,
," l' i
i i"
prey I:txpectl:td
Secondary hiomaS8
(kglkm'l Habita'
Norlnern or
Boreal foresl
Mixed broadleal
Prairie grassland
Aroaca saVll.nna
Rain foresl
The Iransport of anUre bodies 01 bolh large snd smaU
carnivoras jnto the cave or the killing of them while
they wore marauding the cave no! only for lhe
greot number of Iheir faetal remllins (especially Ihose of
tba hyaena) bu! also for Ihe highly elOdedcondillon of ji
large number olbone flakes and deciduous teeth [Kilch-
ng 1963:191
AH of these arguments regarding the subsistence
base oC the early hominids are based 00 nfereneas
frum 8ssociations bctween bones and stone tooh at
alleged living sites. To what degree may this as-
sumption be made casnally and how kely is such
an assumption about the past to represent Ihe past
How "realistic" is Ihe assumplion that as-
sociations between stone t001s and bones at a site
imply beltavioral io:tegratioo: in the pasl?
Given that predation and death are normal eVllnts
in any environment. to what extent may Ihe ar-
chaeologist expect to find evidence of such events in
situations favoring preservation ofbones? This ques-
tion 18being invesligated by a number ofresearchers.
particularly in Africa. To appreciate the eharacter of
this problem, one mus! become familiar wHh sorne
basic faels of nature, For inslance, George B. Schaller
(1972:396) estlmates thll pred<ltors (lion, hyana,
leopard, cheelah, and wild in the Serengeti-
Ngorongoro sludy area "may actually aceount for
exteneon of the monument concept to hebttatcns
and later lo living floors has in my opinion resulted
in numerous myths regerdtng man'a encent pest.
"Tools" have been described thal were probably
producod by hyanas, cave bears, and cther predatnrs,
and the buntng of such unllkely foods as bvena and
cave beer has been cossktered a regular pert of rnan's
The most ncredtble exampla of this Iype of myth
making must be the analysis of materials from Pin
Hale Cave at Craswell Crags, Derbyshire, which was
excavated by Armslrong and studied by Kitching
(1963). There Kitching worked with 6985 identifi-
able bone elements, of which 3197 (46%) were from
carnivores and 1061 were from the hyena alone.
Mixed among cArnivore-dominaled faunal as-
semblage were sorne telltale stone tools. These were
sufficient to lead Kilching to the interpretalion that
everylhing found in Pin Hole WaSin fad the result of
hominid actlvity:
F. Clark Howell searehed for living f100rs al Isimila
and lalar al Torralbe, and tho Leakeys seBrched for
such floors al Olduvai Gorge. In Ibe laller case a new
definilion of a living floor was offered; "Ihe occupa-
tiOD debris is found on a palaeosol or old land sur-
lace wilh a vertical distribution of only a lew im.:hes
[M. D. Leakey 1911: 258J." This definitioo is very
clase lo that which mosl areha8ologists would offer
fOl Eln archaeological aggregation of artifacts
(rellcs) on a and surface! The problem, which
should be clear, is Ihe degree to which the assump-
tions governing "interpretation" of a monument are
justified with raspee! to such pllu:es as living floors
or even habitations as defined by Paleolithic ar-
chaeologists. 11 will be recalled Ihat a monumenl is
considered a construction produced by man; hence,
all the elements, or of the constnwtion
Can be assumed to be there "by Ihe hand of man,"
The interprelative justification Ihal went wilh Ihe
Tha floors are only one mone tbid. snd every ,tone is
eilher an artiloc\, Oake, or anvil. The enclmlng sedi-
ment, are non-pebbly and on Ihe floors thamselves
there are no pebbltls whkh were transported by natural
means; they were aU left in their presenl positions by
prehisloric mano They are shorp and unworn. nor were
Ihey ltpparently moved slgnificanlly from the original
positions b Ihe a;tion of water They were facto!)'
floors in the best sense al lhe tenn (Bond 1969'.2071
and it was apparen! that Ihl! Jilhic and other material
Iying on them representad debris from human occupe-
tcn sitas. developed, therefore, a tedmique for ex-
cavating tbese "Iloors,' end thelater dislurbad surfaces.
off the ovarlyi"$! stertle senos and days
and expostng the fluor, or secton of oor. so that the
relatlonshlp of everylhing on it couid be cleerly seen
and plotted. In ths way we were eble lo obtem detalls of
thp. relationships of the tools. factory waste, natural
stcnee. and wood IlIJId so 01 relatiunships between
classes 01 snaped tcols. thus supplementing fu:! inlor-
matten te be obtatned from a sludy 01Ihe artilacts Ihem-
selves 11. o. Clark 1969:11-121
111 Clark's early work geological crileria and facts
of site structure such as lhe association of debitage
with COTes and anvils appear to hAvehHln the criteria
for recognizing a living floor. This approaeh to
livingfloor recognition is well iIlustrated by Ceof-
frey Band's description of these flaors:
Olorgesailie wa8 8uilable as tesling groulld bttI;aU5tl
arUfac!s could be recovered from fine-grained sedi-
ments Ihat had praserved detalled evidences of ar-
chaeological assocjations. These could be related lo
stratigrapby and to festures 01the original envirQnmen!.
Thls preservalion was in marked contras! lo Ihe sites
that hlld yielded most 01 the evidence on whlch the
c1llSslc interpretations of Middle Plelstocene culture
had prevlously
The lower horizons containlng the artilacts showed lil-
tle evidence Dfdislurbsnce bywaler action burial
Despite the facl Ihat Lonis Lea.key reported the
Olorgesailie living floors (L. S. B. Lee.key 1946),
sorne believe that he did not "realize their mpIILa-
fons" (Vincent 1918:32), as he viewed Ihem as use
fui for assemblage definition and not so much as
.!nueces of behavioral information about Ihe pas!.
The view that Ihe Leakeys treated the 5urfaces more
as "curiosities" is perhaps supported by referenee lo
Isaac's descriptioos of their excavations at
Olorg8sIlIille {1971:2879), which cite pltlfully few
records macle by tha Leakeys for excavatons carried
out between 1943 and 1941. \-iore recent work and
8Ctual analY8ls of !he racts from OJorgesaiHe render
lhe identificalion of living f100rs lesll and les3 Iikely
(see L. R. Binford 1971b: Isaac 1977). Nevertheless.
the search for living floors was first prompted by Ihe
Leakeys' claims at Olorgesailie. Most dearly iI was
in the work of J. Desmond Clark al Kalambo Falls
where Ihe techniques of recovery that have fortu-
nately become somewhat standard in Paleolithic re-
sea.rch were originAlly pioneered. Clark states,
floors was the contributioo of Mary Leekey. who hed
hutaperience wilh ','Neolithic" excavations and an
tnterest in the use of archaeologtcel techniques to
nvesteete the "place," rether Ihan just using the
place lo recover arufects. Severa! persons have indi-
ceted that the Leekeys' original clams for living
0001'5 al Olcrgesetlte were besed en the eegument
that the sedimenta were fine and hence the geologt-
. -cel processes resuhing in burial were "gentle."
Under aucb candillons ODe could expect ttems lo
remaio where they were deposited by agents such as
thIB homlnids of early time ranges. This "origi_
nal" argument regarding living floofs is al least par-
tially presented by lsaBC;
J. Hehes lo Artifoels ond MrmlJments lo Chonging Concepmu! Fmrneworks



, 11

This is a truly remarkable set of conclusions lo draw
from the Torralba dala. We are told that the region is
one 01 "undoubled natural wealth" (presumably
weallh in animalsJ, yet we must assume that none
ever died trom predation or natural causes aronnd
the Torralba lake.bog! We are told Ihat nalure oper-
ated only when man entered the picture. When man
arrived to hunI enimals cooperatively in large
groups. nature also cooperated by preserving a re-
cord of these heroic exploits. Yel, when man left.
nature stopped and no further preservation, or
natural dealhs of Bnimals, occurred! How extraordi-
nary! We are led to conclude thBt. becl,msethere were
so many animals represented in "Ievels" with lools,
human groups were practicing cooperative hunting
drives in which all species over certain sizes were
taken time and again in roughly aqual proportions,
butehered, and lhe meat transported to consumers
living in large groups that were "a conslanl feature
01 the landscape." AmBzing!
The researchers at Torralba have operaled with a
set 01 assumptions abaut Ihe nature 01 Ihe deposits.
as well as abaut the processes in the pasl that re-
sulted in deposit lormation. They have been mis-
guided by the idea 01a living Iloor as a spatial aggre-
gaton 01preserved maller, all of which results from
human aclions. Given an aggregalon 01 slone
tools-evidence 01hominid behavior-it is assumed
that al! olher remains associated wilh the stone tools
are also a by-product of human behavior. The re-
searchers of Torralba have certainly made Ihis as-
300 loeds would have been required to move al! the
meat: 50 able-bodted men would have lo make 6 Irips
eech lo carry the 101 to the living site. The amount of
meat aclually utlhzsd might heve been constderably
smaller than Ihe total available, but lhe figures sttll
suggest a relatlvelyhigh group size. Basedon suchcon-
sideralions, I em inclined lo thnk Ihat the total sree of
the cooperatng social groups which provided the per-
sonnel responslblefor the TcrralbaoccupaUon resldues
wes very large, perhaps on the arder of a hundred tnd-
vduals al more. Such large populalion eggregetes
might hava been feaslble only perlodically 01 seeson.
ally, bul it is quite possible, given the undoubled
nalural wealth of Ihe region in 66rly mid-Pl6islocene
times, thal large human group, were a constant feature
of Ihe landscape (Freeman 1975:682J.
The grealasl estimaled minimum number of animals
for any level is only 15 (Occupation 1). Varialions in
speciescompositiondo oeeur fromlevelto level.bul are
no grealer Ihan can be expacled from sampling error
alone. The mosl abundan. remains are Ihose of
elephants [31 pereenl of individualsJ, allhough lheyare
alwoysoulnumberedby Ihe combined10101 of individu_
als of other spacies.... Horses are next most abUndanl
[21 percent of individualsJ, followedby cervids (18per-
cenl excluding Dama), bovids (13 pereenl), Dama (6
percenl), rhino (4 pereenl), and birds (3 pereentJ, wilh
abouI 2 percenl each of Felis and olher carnivoras
[Freeman 1975:6681.
What role did man (represented by 611 tools depos-
ite<! during "several tens of thousands 01 years")
play in this Pleistocene biotic communily? The re-
searchers' answer is Ihat he contro!led it completely,
dominaling the enlire situation so that every animal
found in the Torralba deposit {with the possible ex.
ception of several birdsl is assumed to have gotten
there by the hand ofman! (See Freeman 1975:667-
682.) Man killed the animals while execuling game
drives-possibly aided by fire-butchered Ihem, and
carried the meat away-truly extraordinary1
Even Ihough much meal mighl hava heen wasled, an
impressive quanli!y must have been transported from
lhe sile, lo judge by Ihe missingskeletolparls. In Occu_
pation 7 alone: animals whose carClIsses wou(d have
yieldedat leasl 30.000pounds of usable f1esh were re-
covared. Assumin!l a daily cons\lmplion of 5 pounds
perman day. Ihis wpreseots about fi(mll man-oaysof
meal rations. lusl bl1aring aWllY so m\lch wl1ight would
haveprl1senleela smal! !lroupwith SeVere (ugistil: prob_
tems. IflJHch lJ8rSllll carrimlllwlIY 100pouuel.s al il loAd.
The rliflJct cnd AssembJage Phuse
land surfeces were dilferentially stabiltzed. permt,
ng the excavators to recognize sequential depo-
stnonal situations. No estimales as to the length 01
time involved in each "stabilization" have been of-
fered. For the entre sequence representing perhaps
"severa! tens 01 thousands 01 yeers" an esttmeted
minimum o 115 (See Howell 1966:121-122) indi-
vidual animals were recovered from the Torralba
levels. Assuming that only 10,000 years ts involved,
there would be an average 01 only one dead animal
every 87 years. That does not appear excessive as a
normal death rale far a normal glacial environmenl!
The fauna is interesting in ils composition, in thal it
is remarkably stable from level lo level:
margfn deposits and interna! dratnage pockets can
be expected 10 vteld considerable quantltes 01dead
animal remains as a normal condftton. Stated
another way. the normal Iuncuontng of a biotc
cornrnuntty can be expeeted lo yield considerable
deposite 01 animal bonas in geomorphological set-
tings conductve to lheir preserveon. This seems
like a reesoneble proposition; nevertheless. the
rnethodological tmpltcetons bave nol been explcred
bv arcbeeologtsts. In Ihe New World many death
ses 01 Pletstocene elephanls have been recorded
and only a few have yielded any evidence 01human
involvement al the time of dealh. The vasl majority
of these elephallts have been recorded in the con-
texts of lake, stream, and spring margin deposits. In
fact. this is such a regular patlern Ihat it is quite
Iikely that elephants of all types seek bodies of waler
when their lemperatures go up; hence the most
I;Ommon place of death is in such a marsh-bog set-
ling. This is a normal condBion.
The eonditions favoring preservation must be
also presenl ifbones depositad as a resull 01hominid
aetions are to be preserved. lf Ihere is evidence 01
hominid behavior in a deposit containing preserved
bones, the conditions under which lhe deposit was
lormed are also Iikely lo be conditions conducive to
the preservalion 01 bones derived from the normal
patlem o predator-prey interachon as well as the
normal mortality characteristic 01the setting.
Perhaps this point is best made with an exarnple:
The existence 01(o) an active biological community
with normal death and predafion rates and (b) a lake
margin hag deposit is well documented at the la-
rnous site 01 Torralba in Spain {see Howell 1966}.
Witbin tbis site were approxirnately 785 "artilacts,"
01 which approximately 174 are lithic debris. These
provide exciting evidence lar the participation of
hominids in the Middle Pleistocene biotic commu-
nity, a glimpse 01 which is preserved at Ihe site.
Tools at Torralba are restricted to Ihe importanl de
posit designated Hc 01the "lower grey colluvium." 1
can find no estimates by the researcher as to the
length 01 lime during which this deposit accumu-
lated. 1 asked HowelI, and he said that he had no
good idea bul that it could nol be more than "several
tens of thousands of years." During the unknown
span 01 accumulation, at least seven "Ievels" were
generated with several additional lenses. Thal is,
boreal forest and a temperare grassland [155.34 kgl
km'). If a depostt eccumuleted over epproxrnetely
10,000 yeers (not an unreesonable estmate for many
Pletstocene events). we would expect 155,344 kg of
prey during thet penod. Given en esttmeted mean
weight o 222 lb (100.57 kg) for 1)0 edult cenbou
(Rangifer tumndus}, we would thereore expect 1544
individual cebou lo heve been depositad over 8
single square kilometer durtng that peeod o time, if
ceribou were the only prey animals in thet babttet.
Obvtously, that is not reelsttc. since many different
spedes would be nvolved. Neverlheless, 8S en oh
ject lesson tor Ihe archaeologist, ODe can sea that an
impressive amount 01 bones is deposited annuall:v,
particularly in a tropical savanna. Given the Iypes 01
geological units archaeologists work with in Pleis-
tocene deposits, we can expect as normal a very
large numbar 01included bones, since most geologi-
cal deposits represent appreciable periods 01time.
Certainly it must be realized that aH places where
deaths occur are not equally propitious lor the pres-
ervation 01 bone. Recent pioneering work in
taphonomy, partieularly by Diane Gifford, K.
Behrensmeyer, and A. Hm, has bagun to illuminate
the conditioos under whlch preservation is most apt
to occur; a quick burial, and either the perpetuation
01 moisture on the bone, or complete desiccation.
Stated another way, in situations in which it is per-
lectly wet, or perfectly dry, and the bone is being
bnried, bones are most apt lo be preserved. For aH
practical purposes this means that areas being de-
flated (see Gifford 1977;108-257) or eroded are not
good candidates, and areas subject to extremes 01
wet and dry are not good candidates. Lake and
slream margins, delta lormations, as well as pro-
lected areas such as caves and rockshelters, are very
likely placas. Behrensmeyer (1975b) studied bone
distribution and assemblage contenl in several dif
lerent geological contexts. She observed lhat bcme
densities were greatest in della environments and
least on the floodplain, with channel deposits inter-
mediate. The implications of these observations are
Ihat (o) death and predalion result in considerable
numbers of animal carcasses beiog scatlered in the
envirooment annually, and (bl these carcasses are
apt to be preserved in dry or wel conditions where
the bones are apt to be buried shortly after deposi-
tion. This means that lake, spring, and streBm marsh
1. Relics to t\rtifacls and Mormments lo Assemblcges: Chonging Conceptual FrameworkB

The Artifact and Assemblage Phase

Pletetocene archeeologtsts need to ebendon such
en approach. They must adopt an epproech that
vrew deposils yielding stone tools 88 natural depos-
ts. recognizing the potenttal sccumuleucn of de-
rivativas from a cross section of biotic adivity in be
environment o a loceon favoring preservatioo.
Man is only ODe potentiaJ contnbutor.
Mydiscussion up to this poiot has focused 00 Old
World metertels and 00 Peleoltbtc research. 1do not
wish lo imply that rny crttctsrn s restrir:tad lo such
studies. In the New world, argumente over the stg-
nificance of bone modifications are relatively recent 1
It is true that Webb 8nd DeJarnette {t942:763)
suggested, in their "generalized profile" of tbe Flint
ereek sbell mound alang tbe Termessee River, Ihal
fuere was sn early period characteri:r.ed by worked
bone. followed by a Iitbic periad, aud finally by a
ceramic periodo However, tbey did not imply tbat
bone was the exclusive raw material used in tool
manufacture. The latter position, linked with same-
times remarkable daims for Ihe antiquily of man in
North Amenea, is a pasition lhat has developed rela
tively recently stemming from work at Old Crow
Flats (see Harington et 01. 1975) and excavations by
Dennis Stanford (1979b) at Ihe Selby and Dutton
sites in Colorado. ISee Canby (1979) for a popular
discussion of this materiaL) In addition to the "os-
leodontokeratic" claims for man in North America
perhaps as early as 60,000 years ago (see Canby 1979:
348; Friedman 1980:7). Ihere has been a stepped-up
sel of claims regarding the use of bones as expedient
lools for butchering. This argument was most force-
fully set forth by George Frison (1970) with regard to
materals recovered from the Glenrock Buffalo Jump
in Wyoming. Since that time this argument has cer
tainly captured the imaginalion of a number of
worken, snd excavalors are now increasingly "see-
ing" bone tools in Iheir faunal assemblages (sea, for
example, lohnson 1978).
Slone tools are exciling traces of homind parlici-
pation in an environment, bUI they in no way may be
lreated as prima fade evidence hat human behavior
is responsible for the deposil in which they llCcur.
This is not to say that habitations or living floors do
not exis!. They certainly do, and as we kllOw from
'he archaeological record of relatively recent times
IAnexception IS Merriam. 1906 llnd Pulnam, 190fi.
human behavior mey well be, on occasron. the agent
reeponstble for the locauon and conlent of meterals
eematnng from the past. However, this is somethlng
that requlres demonstreton. not something lo be as-
We need to ebandon the logc of treating alleged
hebtatons. living oors, and living stes as if they
were monuments. The 10Mic is one ot identifying an
envelope or container. and then uncritically accept-
ing everythng wthn t as referable to the behavior
o encent mano A complete reevaluaton of besc ar-
cheeologtcal concepts and the logle associated with
their use is in arder.
Minimally. let us begin by asserting that an oro
choeologicol faet is an accurate observation tha\ caro
ries unambiguous information relative to hUIDan-
hominid behavior. The degree lo whkh an associa-
tion between two or more Ihinga can be taken as an
unambiguous archaeological fad is alleast partial1y
conditioned by the character of lhe deposit; that is,
by the matrix or "container" 'hat bounds the associa
lion. Most would agree that archaeological remains
are items and arrangemenls of maner thal are de-
rived from Ihe operation oC a human-hominid sys
tem of adaptalion. An association between e\ements
im::luded in a deposit may be uncritically accepted as
archaeological fad only when the deposit or matrix
containing Ihe elements in association is Iself an
archaeological deposit, and man can be viewed as
the egent responsible Car the presence oC both Ihe
associations and elements wilhin the deposito
ArchaeoJogicl1/ deposits are matrices \oMelher
with ncluded remains thal are derivad from Ihe op-
eralion of systems of adaptation.
That is, the matrix itself is an archaeolop;ical phe-
nomenon. From sites of relatively complex cultural
systems. archaeological deposits are common; mid-
den accumulations, mounds composed of collapsed
building materid, and mounds specifically con-
slrucled for the burial ol the dead are a few exam-
pies. Archaealogical deposits may be expected to be
Tare and infrequent derivalives of hunler-gatherer
adaptalions, since high mobilil)l and relatively low
investment in facilities, housing, buildings. and so
forth are characteristic of such syslems. It is much
more Iikely that archaeological remains will be
found wilhin geologicaJ deposils. Geological depos-
its are buried matrices, Ihe formation of which de
nves from the operaton o geologtcal processes. Ma-
tertels wtthn a geologcal deposlt may derive from a
wide variety of circumslances and events that oc-
currad during the deposu's formation, none ofwhich
is necessanly representative of a behevorally nter-
related set of condutons. Thal s. en cssoctcon be-
tween two or more elements within a geologtcel
depostt in no way mples a systemeto. behavorel.
or dynamic relationship between the componente.
AIl that is mpl ed ISa roughly coincidental contex-
tual releucnshp for the events of which the ele-
ments were derivalives. For nstance, the finding of a
whale bone and a luna vertebra in a gaological depos.
it in no way implies that whales ate tuna or for that
maltar that luna ate wheles' Nevertheless, the depo-
silional contexl of an aquatic environment may be
indicated. Geological deposits may be thought af as a
kind of grab bag. made up oC samples of byproducts
of events and enlities that existed or occurred with a
given "catchmenl" and during a given period oftima
when tha deposit was accumulating. A good exam
pIe is the preserved land surface at Laetoli where
dampened volcanic ash hardened to preserve
hominid footprinls, depicting a Slral1 by two "early
men" across a \and surrace sorne 3.5 million years
ago (see M. l..eakey 1979). Also preserved in the
"ashen ledger" were trads of elephants, guinea
fowl, giraffes. hares, ostriches, and nther animals
and insects, as well as planl remains. Are we to as-
sume thal Ihe association of hominid footprinls on
the sama land surface wilh those of other animals
indicates Ihal the action of the hominidsconditioned
both the presence and charader oC the patterning
observable in the other foolprints? Was early "man"
herding elephants, ostriches, or guinea fowl? This is
clearly an absurd question. The footprints and olher
traces of pasl events and aclions must be viewed as
roughly contemporary but unrelated in any inte-
grated senSe. That is, as far as mulual interachons
and "causal"linkages among the events recorded are
concarned. none seems indicated. Clearly more jus.
tificalion for assuming a behavioral integration
among materials associated in a geological deposil is
nseded than Ihe simple assm;iation itsc1f.t
As a way of placing emphasis on this fact, it is
perhaps uSeful lo think about aspecls of pRsl pro-
cesses or r:ondilions Ihal might result in different
kimls of geoJogicfll neposils. For instance. it is
perhaps prottable to ask oureelves about the hstcri-
cal integrity and resolution of eggregetes of matertals
eteblteed in depositional aseoctetton throcgh the
operaton of gcologlcal processes.
Historical intesrity refers to the degree lo which
nclusons wlthin the depos derved from the same
or dfferent dynamic conditions in the pest. For in-
stance, if all the materials in a depoet derivad from
the aelion of homtnds. we could argue thal the de-
post had considerable integrity. Sirnilarly, if all the
Included material wes referable to the actions of
hyenas at a den, predators in killing prey. or the
natural deaths of animals. each would command a
judgment Ihat the integrity of the deposit was grea\.
On the other hand. if all the mentioned agents were
represented in a deposit, we would say the deposit
had liUle historical integrify.
The relative resoJution of an accumulation of ma-
teriaIs is to the degree to which items and dasses of
malerals may be referable lo a specifiable and lim-
ited, hence unambiguous, set oC events or acUons in
the past. Assemblages with high resolution are as
semblages in which all parls are referable to \he set
of events or condihons in Ihe pas!. Resolution of as-
semblages may vary independently of the degree of
ntegrUy. For instance, we mighl have a deposit with
high integrity in thal all lhe included materials are
referable to a single agent-hominids. Yet the events
and activities that the hominids participated in
might span a considerable period of lime and repre-
sent a wide range of different behaviors, which have
in eommon only the facl that they occurred in the
same place. l Ihink it should be dear lhat ltIJtjjrify
refers to tbe,relal.ivehomogen&ity "rthft. ogeRta-re-
sponsible for d nlu-
tion reters to the relativa homogeneity oj the.events
of situationaJ condition8 whose- by-produetnn'e'pre-
served .in.the-depbsil.
The basic idea as to the relative resolution of a
population of depositionally associated items has
been conceptualized previously. Archaeologists
have recognized that deposits yielding archaeologi-
cal remains might be multicomponent in character,
in lhat the deposits derive from the independenl oc-
cupalion ofthe location by popula1ions represenling
differenl cultures, time periods, Of his\orical eras.
Thus. archaeoJogists could speak. of a site having
severnl occupational {:omponents.
: 1'1
11 .
: ' '


1. ReJies lo Arlifacts ond Manuments to Assembloges: Chonging Conceptual Fromewarks
In treattng sttee thet were occupationally dif-
ferentiated within a system. 1adoptad the term grcn
(L. R. Binford 1978h:482-483) fa refer fa propertes
o assemblages. Ftne-grcrned assemblages are those
in which all the included tems. features, and land
surfeces relate lo a very few aventa; that s, all as-
sociated archaeologtcel charaderistics of the deposit
are the ennsequences of basically the same events.
Coorse-grained ossemblages accumulate over a con-
o -atdereble period o time andJor durng pertods o
rapid "tumover" of events. resulting in the cssccrc-
neo of ttems, debris, eeturee, land surfeces. end the
ltke thet were differential partcpants in different
events during the course ot the occupetton. Thus, the
tenn grojo refers to the relativa contextue! complex-
ity of en assemblage from the perapecuve of events
occumng durmg the course of a continuous occupa-
tion and derivativa praduction of en archaeologcal
essemblage. Recognizing that deposits may be vari-
able in integrity, and in turn in resolutton even given
high levels of Integrlty, we must minimoJJy have
sorne releble means or referring observed palterns of
association to potenttally different formation con-
teds. We must further recognize that the resolution
of an assemblage must be assessed at !reveral organi-
zational levels befare meaningful comparisons con
be carried out. We may find assemblages lo have low
resoluUon because of independent oceupational epi-
sodes. or beeause of Ihe relative redundaney in the
evenls oe"urring within a given occupalional epi-
sede. These and other derences surely contrtbute
to differences in content and hence meaning.
In later chapters of this book it will be em-
phasized thet cerntvoree are apt lo generete dsttnc-
ttve faunal assemblages in basically Iwo oontexts: at
kills. and wilhin and around lairs. These Iwo types
of assembiage. as well as contributions from natural
deaths. can be considerad as very likely contrtbutors
lo the "background" faunal materials occurnng in
most any geological depostt where bones have been
preserved. As suggested eerlter. preservatlon is most
likely in caves and rockshelters and in open deposits
aceumulated where burtal was relauvely quick and
where mosture was ether very hgh or very low.
I have suggested that most f not all. locations
yielding evdence o cur Pleistocene anceslors are
most likely geolcgcal deposits, not archaeologtcal
deposts as IS commonly assumed.
This assessment of the character of our Lower and
Middle Paleoltthc data demands thet we develop
means for recognizing the dervattves ol differenl
agente and differenl events as conlributors lo the
geologlcally associated palmpsests withn which
may occur sorne traces of hormnd bebavtor. We can
no longer be conlent wilh such tautological condi-
Iions as identifying a habilation or living site by
virtue of the presence of reHcs and ather associaled
remains-generally bones--and then using the dala
from such sites lo prove the nalure of man's horne
life or other c.haracteristic behaviors!
In the prevtous chapter t showed thal the essump-
lions made regarding the condilions under which
the archaeologicai record was formed dtrectly cond-
non the charaeter of inferences about Ihe contenis of
Ihe archaeological record. I showed Ihal we may be
frequently incorrect or alleasl highly uncertain about
our reconstruclion of the pasto In Ihis chapler I will
explore a somewhal more complicaled issue---how
we might proeeed so as lo minimize Ihe likelihood of
conslructing false piclures of Ihe pasto 1 will be di-
rectly conr.ernerl wilh fflsfJarch lar.tics and how we
mighl use secure knowledge lo aid in Ihe develop-
ment of new knowledge or understanding regarding
Ihe past. How do we carve out knowledge {rom igno-
The challenge lo archaeologisls is simply Ihis:
Haw do we proceed? How do we unify Ihe world of
archaeologicallhings wilh OUT ideas as lo Ihe charac-
ter of Ihe pasl? Haw may we use Ihe empirical world
of archaeological phenomen() lo slimulale ideas
aboUllhe past and at Ihe same lime use Ihese empiri-
cal experiences fo evaluale Ihe resulliflg ideas? How
can WfJ proceed so as lo develop confidence Ihat OUT
ideas I}f Ihe past are informative abolll the aclllal
pasl? Wl' flH;C Ihc challenge of scitmce ilself-how lo
kecp our fep.t on !he "empirical' grounrl <lnd our
hearlgin Ihe 'theoretical' sky. Basic lo !he df!\'nlo)l-
mf'nt (lf a is a of dOBl<lill (o
Chapter 2
research and the
role of actualistic
I i
2. MiddJe-Ronge Hesenrch und the Role 01 Actunhstic Studies

The P{]radigm-ne's Cuide lo Di!srribing the World
which scientific procederes might be profitably
Mkkessed--empirical with respect lo what? Theoret-
ice! with reepect to whetv
Many ercbeeologtsts accept the argument that the
discipline of archaeology needs to edopt a sctenttc
approach, yet they are not ecessertly in agreement
as to the domatn of expertence to which such an
approach is to be rnost profitably addressed:
there s in a aense en "archaeologfcal theory" although
it mighl be better characterized as evoJutionary an-
thropuiogy ... human sud cultural evoluon is o such
scienlific and intrinsic nterest thBt here is certainly an
"sential nomolhelic role lo be piayad by an::haeologists
[W8tson el 01. 1971:1641
In the foregoing view, archaeological theory arl-
dresses a dornaio of pest eVents sud conditions. It is
concerned with explainiog why certain evenls and
syslems carne into beiog in Ihe pasl. Haddresses the
dornain that most Iraditional archaeologisls consid-
ered lo be their target or seeking understanding:
such interesting problems as the origins o the slate,
the shift to agrieultural production, or perhaps the
origins o culture itsel. Under this view 01"doing"
archaeology, Ihe scl of investigating the archaeolog-
ical record is viewed as the experimental phase or
perhaps the archival phase o invesligating Ihe past.
I have referred to such inleres!s as general research
or general Iheory building. By Ihese phrases 1refer to
Ihe actions of investigators seeking to explain
characteristics of Gultural systems past and presento
The domain of nterest is cultural systems. how they
vary, and how they may be modified from one form
to another. The domain is interaclive. generative.
and dynamic.
Important, however. is the ael that all knowledge
of Ihe dynamics of Ihe past must be inferred:
To say Ihal historillns conslruct the pa.st so as nol lo
falsifyr:erlainIheorelical presupposilions is no!to point
to a derecl in historians or inlheir method. Jt is lo focus
upan the mellns histarians IISe lo finrl out what hap_
pened. [1 is lo sayo in a somewhal di(ferent way. what
has been said berare: !ho! wr. the even!s of !he pnsl
fromlhe evenls lIt 'he presenl by linking them in
of sorne Refiero! printiplf's IKitts 1977,1)7 6f11
If we recognizc lnal science is r.oncerncd wilh
cicveloping means fur im:reasing Ollr undcrstancing
or observations of nature. Ihis mplies that sctence is
stmulteneously ettempttng to generate understand-
ng and to sherpen or lncrease the ntormattonal po-
tenttal of our cbservatons. The archaeologtst inves-
tgates pbenomene that he has reason lo beleve re-
mein' from the past. These lnvestigations are con-
ducted in the present. resulttng in all the observa-
tional steternents genereted by arcbeeologtsts being
contemporary Iacts. How does Ihe erchaeologtst
conven these eontemporary observational stete-
ments or faets into meaningul statements about the
past? The first thing that must be realizad is that Ihis
can only be accomplisbed intelleetually or with ree-
son. Thus there is no way of converting observa-
tional statements aboul Ihe present into meaningfnl
statements abont the pasl in the absence of a reason-
ing process.
Insofar as archaeology remains a discipline that
searehes for an understanding of the past through
the use of objects and olher organizatiom; of matter
believed lo have been parls of pasl stuations, ar-
chaeologisls musl operate as historians attempting to
give meaning to observalions on Ihe particular ar-
chaeological record being investigated. The accu-
racy with which we may accomplish Ihe conversion
of contemporary observalional stalements into
meaningful statements about Ihe pest is a direct
funcHon of Ihe character of Ihe reasoning processes
employed and Ihe methodology developed for
evaluating the products of Ihis process.
Our problem is then twofold: (01 We musl know
the past by virtue of inferences drawn from knowl-
edge of how Ihe contemporary world works. the
principies mentioned by Kilts, and (b) we must be
able lo justify the assumption thal Ihese principies
are relevanl-Ihat al leasl in lerms of Ihe properties
of Ihe principies, Ihe past was ke the present; we
musl make a uniformilarian assumption.
The poinl Ihal we must use general principIes in
giving "historical" meaning lo our observalions no
longer seems al issue:
rrom Ihe generalizlng social sciences
only in Ihal ils primary aim is lo eKphlin individual
situallons in a1l Iheir comploxilY r81her lhan lo farmu-
latA gAnerallaws fDT indf'firtelyrepealable evenls and
prm;esses. Tha! ls whlllis mtlanlbysaying thal hislory is
idiographk. Ihe s(l(:ial science nomolnelil: INall'll1
1%1:547; Elto/l 41). This does nDI mean
thal hislorians rleny Ih", exiS!f'm;e of genr.ral rulAS:
rather Ihey seek to employIhem lo gain an undarstand-
ng of individual (Le.. unique] and non-recurren srrue-
tiuns [Trtggar 1978:26-271.
where do such general prtncples come from, and
how can we be assured of Iheir accuracy and
vanee lo our acfivifies as archaeologists seeking lo
explicate the rast?
Those who daim that archaeoJogisls should be
htstortans and not scentsts most commonlv edvtse
that we should "borrow" our general principles from
other nomothetln sciences. The lrouble with this
suggeslion is that I know of no nomothetic science
aftempling to undersland lhe archaeological record!
Many olher "sciences" may be concerned wilh vari-
ous aspects of human behavior, history. and
sociocullural change in which Ihe phenomena stud-
ied are events, behavior, or palteming in communi-
caled Ihoughl. However, Ihe basic phenomena with
which we work are (al static, lb) material, and (e)
untranslaled into symbols or c1ues to human
"thoughts." No other "science" addresses such
phenomena. It was Ihe recognition of this fact thal
prompted the following stalement wrilten in 1966
and published Iwo years laler:
Accepling Spaulding's minima! derinition of what ar-
chaeology is. we can go a slap rurther and specify its
aim as the explanalion .. of thE! order we observein !he
archaeological record. AlT.haeologlcal theory consist.'! of
proposilions and regllrding the 8r-
ehaeologkal Tlword ilself-Us origins. ils .'laurees ofvllr-
iability, Ihe delerminants oI differences and similarHies
in the fnrmlll, spal\11. and lemporal chllraclerislles of
arlifllds and featur..s and their inteITtlllllionsnlps lS. R.
lJinfordand L. R. Hinford 19611:21
Directing altention lo Ihe archaeological record
ralher Ihan continuing Ihe self-deceit Ihal we were
sludying Ihe past seemed e:enlral to progress. My
view was Iha! we could not reconslrucl hisloey unll
we fin;1addresscd the problem or how we give mean-
ing lo the arcnaeologkal record [sce L. R. Binford
1968d). Meanings are carried by concepts and rgu-
menls and the arr:hapological record contains only
llrrangenllJnls of mfltlN If flrchaeologists are lo know
lInything of the PflSI. Ihey musl dcvelop a science.
The r10main of this Scif'llce musl be the archacologi-
cal Hlnrci pr't !w
In seektng lo develop a science o the ercbeeclog-
ical record, are there not sorne fundamental charac-
tersttcs of bcth scence and the archaeologcet re-
cord that we must consider to guido the growth of
this scencev The answer musl of course be yes. One
cherectenstc particulerly lrnportant lo the argu-
mente advocating a science of the archeeologtcal re-
cord is that science attempts to eveluete the role and
utility of ideas for enhanclng understendng. Ideas
are of course cultural forms:
if we view culture as er least referring lo tbe parttcularly
human abilily lo gtve meenlng expediently to axpen-
enee, 10symbol, and in turno view experience through
thls conceplual ldiom. science is then eoneemed wilh
evaluatlng Ihe ulilily of he cullural lools produeed lL
R. Binford 1977a:31.
The reference to Ihe "cultural lools" produced is
of course to Ihe concepts and ideas in terros of which
we conceive Ihe world of experience. If we gain a
"knowledge" of the world Ihrough the use of cogni-
ti\'e devices, words, concepts. and ideas, and the
world is described in these terms, we musl face the
problem of Ihe accmllcy, utilily, and "reality" of
such cognitive devices themselves. This is one fun-
damental problem the scientisl must face. The sec-
ond problem relales lo the degree lo which we seek
knowledge and understanding beyond simple de-
scription. We frequently altempt lo understand why
the world is the way it appears to be, given Ihe
description generaled. Sdentists carry oul their
work wilh essenlially two sets of intelIeclual lools: a
conceptual frame of reference or paradigm (Kuhn
1962). and various Iheories Ihat seek to explain Ihe
world as "known" through the use ofthe paradigm.
The Paradigm-One's Guide to
Describing the Wodd
The cognilive frame of reference or paradigm
r.oflsists of Ihe ideas and concepls with which we
approach experience. These condilion whal one
considcrs relevan! to describe or chooses lo discuss
as of i!lINesl. One's cognitive frame of reference may
he lhoughll!f flS the culture 01' a sdence. It consists of
! I
, I
I :,

2. Middle-.flange Reseorchand Ihe Roleof Actucltsuc Studies
"d".- ..

Theory-c-One's lo Explotning Ihe World

the concepts in tarros of which experlence is in-

tellectually esstmlated. Despte all Ihe defntional
conlroversy (see Masterman 1970]. 1 follow Kuhn
(1977) in viewng paradigm as the tntellectual terms
upcn whch une meets expertence. The cheracter of
one's frame of reerence condttons what te consld-
ered relevan! lo describe, what s interestng to dts-
cuss. and even how we view the world in terms of
problema to be salvad, In short, lt s what we expect
Ihe world to be lke. Things become complceted
when we recogntae that we cannot galn 8 direct
knowledg8 of Ibe essential properties of Ihe world.
Qur cognilion is neilber direct nor objective. bul may
be indirect and subjective relative lo OUT beHefs
abou! the world (i.e., OUT paradigm).
We generally defend our clairns about whal Ihe
world is like with inferential acgurnenls. I prefer to
call these warranfing arguments; Ihey are argumenls
advanced Ihal tend 10 warranl lo others lhe heliefs
one has ahout the world. If done in a robust manner,
they make one's claims appear plausible, and ac-
ceplahle (o alhers. Rarely are such arguments for-
malized in thal Ihe premises are rarely explicilly
stated, so conclusions are warranted hy appeal lo a
"common body of knowledge or belief." The more
comprehensive Ihe alleged knowledge, or wide
spread the helief serving as Ihe nteHectual conlexl
for a warranling argumenl, Ihe more plausible it ap-
pears aud lherefore Ihe greater m.p.lihood it has of
being accepled.
Working wilhin a frame of referenee is similar lo
participation in any olhl"r culture; we aceommodate
experience through our shared cognilive devices.
The facllhal Ihey fadlilah'! Ihis accommodalion ap-
pears lo os as proof lhal the world is in fael lhe way
we expecl il to be. We may be aslonished Ihal olhers
do nol see Ihe world Ihe way we do. Anlhropologisls
should be familiar wilh culfural differences and
should be fairly comforlable wilh the idea Ihal the
nalure of experienee does nol necessarily delermine
Ihe nalure of culture. Many persons share idenlical
experiences yel ascribe lo Ihem very dinerent mean-
ings; Ihis is essenlially Ihe message of anlhrupology.
Cullural man has for alllillle believed thal his beliefs
were given by "reaUly" anrl Wl'lre therefore "righl,"
whereas those of othcr c.ultures werc dearly mis-
guided or "slupirl" {or nol having smJll the "Irulh"
inherenl in giVlTll !1xperience.
Arcbeeclogy is perhaps in a ortunete postton.
Although there ts much contemporary "culture" or
paradtgmauc bias regardng the nalure of man and
the causes of hiatory, there s very ltule folk knowl,
edge regarding tha formalion of the archeeologtcel
record. Thls means thet there is Hule explicit prior
developrnent of cognitive devices and frames of
reterence for eccommodang erchaeologice! phe-
nomsna in the literal, static sense of the word.
For Ihe further develooment of archeeology, the
growlh of a paradigm. developing cognitiva rneans
for idenlifying properties of the pasl or diagoosing
the archaeological record and Ihereby giving mean-
ing lo Ihe archaeological record, is crucial.
Much of Ihe time use of a paradigm is viewed as
an acl of identification. Can we identi(y a habilation,
a hide scraper, a malrilineage, a base comp. agricul-
lure? Or can we diagnose Ihe funclions of a sile, loo!,
or elernent of debris? In most cases we are seekiog an
unamhiguous definilion. and realistic concepts wilh
which lo parlilion or diagoose Ihe archaeologiclIl re-
cord aod thereby generale meaningful statemenls
aboul Ihe pasl. AH such inlerprelalions lIre depen-
denl on a general, accurate, and unambguous
knowledge o( Ihe relatiollship betweeo slatics and
dynamics, the formal consequences for organized
malter Ihal derive from the operalioo of a dynamic
system. In developl'ld sciences, whal is being sought
here at Ihe consclous levellhrough "middle-range"
research may be laken for granled as paradigmatic:
The dislinetion belwpen "empirlcal" and "Iheoreli-
cal" may be only a rellltiveone. It is relative hislori-
eally ... Asclentisl who underlakes the study of a par
licular problem. for example of a biologiCIII oOe. IInd
who uses various scienlific llslruments cOllstru(led un
Ihegrounds of differenl physiCIII theories. s quile aWl:Ire
of !he faellhat logelher wilh Ihe t'quipmenl he uses he
aceepts alsolhese Iheories_ In spile of this fae!.however,
he willlreat the statemenls he will formulale by means
of these instrumeols as ooservlltional. The ubservalional
language is, for him. somelhing already presenl and hls-
torieal1y given hy the developmenl of science and com-
moo know!edge IAmsterdamski 1975:81'\1.
An RQl\eKiS-
tP.Rt:--",,-'::W6h... The cUllep.pls and hnce
paradigmalic characleristics of Iradilional archlle-
olop;y are hp.IiIJVcrl lo be Ilseless fur
rnodern ercheeologv. Today the archaeological re-
cord ls not betng vtewed (by mosl) as a material man-
ifestanon of mental phenomena; l Is 001 being
viewed as a preserved past. il is not beng viewed as
uniquely determinad by history; ts vertebny ts not
beng viewed exclustvely as a manifeslaliun of past
ethntc variability, and so on. As is suggested by
Amslerdamski, the instrurnents that permit and
Factlitete unamblguous meaningfu! observations
musl be developed, demonstrated. and tested, using
sctenttc means. Later. as the scence of archaeology
becomes more malure. lhese "inslrumenls for mea
suremenl" may he laken (or granled and resulls of
their use trealed as direcl observalions on Ihe pas!.
We are a long way from this level of maturily
today. We need lo recognize very explicitly lhe cur-
rent stale o( the arl and address lhe growlh of a new
paradigm as basic and fundaml'lnlaL Recognizing
Ihat this is a historica] phase in the growlh of the
"new archaeology"- I began using a special lerm for
Ihis endeavor: middJe-rangfL resear;h or .!!!idd1e-
range Iheory buildi'!&.l
.. Whal we are se
Ihrough middlerange re-
search are accurale means of identification, and good
inslruments for measuring specified properties of
pesl cultural syslems. We are seeking reliable cogni-
live devices; we are looking for "Rosetta stones" thal
permil Ihe accurale conversion from observalion on
slatics to sh:ltement about dynamics. We are seeking
lo build a paradigmatic frame of refenmce for giving
meaning lo seiected charBclel'islics of Ihe ar
chaeological record through a IheorelicalJy grounded
body of research. rather Ihan accepling folk
knowledge--Iet alone implicit folk knowledge---as
the basis for describing the pas!.
Theory-one's Guide to Explaining the World
Theories ar" the key lo the scielllife underslandin!l uf
empiric:Bl phenomena, and lher are oormally (Ie-
vl'lopeu whl'ln prevjous reseorch has yielded a
'This is l'lssenlially idenlicaJ lo whAI David Clllrke
called IClllrke 1973:8)and appellrslo be
whal Sehiffer (1976) means by behavioru! orcho!'olo;r
(Seealso Sullivan 1978.1
body of infarmalion, induding empirtcu] generaliza-
ons ocout lhe phenomnnrr joqueslion. A theorv rs then
lntended lo provide deeper understandtug by present-
inJol those phanomena as manuestattons of cenan un-
derlying processes [Hernpel 1977:244: ernphasis mine]
Given thal we have made observattons en the ar-
chaeologtcal record, offered sorne generalizacns
about lts pruperlies, and gained considerable experf-
ence with Ihe record, I must now ask Ihe crucial
questton: Why ls the archaeologtcal record the way il
appears lo be? When we seek lo reason about Ihe
"causes" of Ihe world as known, we are atlempting
to build Iheories about Ihe world, "Where it is sorne
evenl or s)'slem of events Ihal is lo be explained,
explanalion has lo do wilh cause [Quine and Ullian
We are concerned wilh organizational properlies
of the world. We seek to undersland how Ihe prop-
erlies of eotilies and/or evenls were produced in
characteristic WBYS:
One very centllll use of "Iheory" involves en epislemic
device which is used lo eharaclerize Ihe state-ehange
behavior of isolaled syslems wilhin H !Ien'lral dass of
phenomena .. one CIIn disr.o,'er lhallhey Ilheoriesl n-
variebly poslulale a c1ass or stales of syslems' change
over lime ... and are used lo characterize how natural
classes of phenomena would behave if isolaled ISuppe
Quite Iilerally, lheories are lhe answers to the
"why" queslions of dynamics. They are concerned
wilh underslanding variabiJity and how syslems
proceed from Olle slale lo anolher.
lf we are going lo build a Iheorelically informed
paradigm for referTng observations 00 lhe ar-
chaeologcal record lo dynamic condilions in Ihe
pasl, where do we begin? It seeros to me we must
begn wilh certaio (undamental slalemfmts of "being
as such_" The archaeoJogir:ai record is a static con-
lemporary phenomenon. It is slruclured malter
molionless and noninleraclive in terms of Ihe prop-
erties of hislorical inleresllo the archaeologist.
Only11univflrseof energy eould have no p8st. flhere is
maller, struclures gro..... and differenliateaod a pasl can
be recognized and rewnslrucled. [1 is Ihe prob-
lem of dllralionless ooll-malter versus enduring mal-
:\\",,:>IUT 11
. U:::='

,1' i'
! 1
(7Z4"1'9z f
Jvl"'l.f? -

2. Mjddle-Rrw8" aesecrcn und the- Rule uf Actunlisnc Studres Theory-c-Gne's Cuide te Explojnjngthe World
ter. .. At ene end of the spedrum is btbltcel chaca. B
pest without 8 pest, beceuse no meuer exists \0 convey
'fnformeon. Al the other end there is only informalion
and no dectsions-c-stenc informal ion Iurever (Margalef
The archaeologist ts o course working with statc
informatioo preserved in structured arrangements of
metter. Since tbere is no energ v ramainng: there are
no culturally relevan! interactiva relalionships lo be
-vbserved in the erchaeologtcal record. SULhrcleton-
shipa existed in the pes! but ceesed when system-
serving energy was no longar powering the rear-
r80gement and modification oC matter-in sharl,
once 8 slatic condilion was achieved. In a very essen-
tial way Ihe contents oC the arehll.eological record
must be viewed as produets of a complex mechani-
cal system of c9usatian. It was mechanical in that lhe
fundamental genesis of lhe archaeological slructure
is a siluation of forces acting lo modify maller in
both its organizational and dislributional properties.
Tha archaeological record is a strudure of relation-
ships between Ihe dislribulion and forro of matter as
caused by energy souoces acting 00 maner in the
pasto In one ver}' important sense, all properties of
matter, whether they be chips removed from a flake
of flint, rniJl;ing of soil betraying the fonner localion
of s pit, pilesof debris from meals, ar the remnants of
a construclion such as a mud brick walJ, are the
mechanical consequenees of the actioos of forces on
term cot!sol in the literal sense, thal is,
lo express the idea of a ealegary of generii::: connec-
tioos; il refers to the way of producing things. Dr.
"something, E, is broughl forth by something else, e,
in a necessary ([;Dnstanl and llnique) manoer lBunge
Clearly if we can solale causal relationships he-
tween things, and if we can understand such rela-
tionships in terms oC more general principies of
necessity, such as the Iheories of mechanics or sorne
other basic science, then we have a slrong warrant
for Ihe inference of the cause from the observed ef-
fects. We would be building a strang Iheoretically
infonned bridge belween properties of Ihe conlem-
porary archaeological record and characleristics of
Ihe dynamic pasto
Insofar as our inferences regarding Ihe pasl fefar
to Ihe causal relafionships that ablained between
dynarncs and tts static derivatives. then any al-
tempts to dtscover the cheracter of such causal rela-
uonshtps rnust reasonably be conducted Ihrough the
study ot living systerns where hcth dynamics and
statc derivativos may be potenlially observed. Tak-
tng as an example the problem outltned in Chapler 1,
the identificalion of the agency [energv sou rce] re-
sponsble for generatjng certain pattams remalning
in the archaeologfcal record. we might reason as fol.
Firsl, we musl attempl lo isolate the differenl
agente or torces that might be expected to contrbute
to or "cause" a gven paltern. Secando we would
have lo conducl sludies of Ihese agenls or processes
in the conlemporary world so as lo develop criteria
of rer.ognition. In shorl. we need lo specify critera
for recognizing lraces, "signalure palterns" apt lo be
preserved in the archaeological record, of Ihe agenls
likely to have conlribuled lo deposils in which
hominid remains migh! also oceur. The procedure is
similar to Ihat painslaldngly worked oul over the
years for recognizing lilhir malerials modified by
man as opposed to slones modified Ihrough olher
natural processes. The problem is one o pattem rec-
ognition linked with the demonslration that Ihe pat-
lem is redundanl and unambiguous, a diagnostic
signature thal discriminales one agent or sel of
agenls from an01her.
8uch a demonslration must be developed by
sludying phenomena generated in a (:on-
lemporary setting, since !here must be liUle problem
of inference regarding the identify of Ihe agent pro-
ducing the pallerning or traces that one is demon-
strating as a signature paHern sufficient for the un-
ambiguous identification of Ihe agen!. The problem
is one similar lo Ihe developmenl of a key identifica-
lion of animals through the sludy of their foolprinls.
The persons who develop Ihe knowledge thal per
mits Ihe recognilion of Ihe track, and nence the iden-
tificalion of Ihe animal responsible. musJ sludy Ihe
footprinls of idenlified animals so thal Ihe relalion-
ship between animal and Irack is a controlled or
known relationship. Given such a control in the con-
temporary world. and given that one is successful in
recognizing and dest.ribing dagnoslit.: criteria (con-
stan! and unique) belween cause and eHecl, animal
ano l"ootprint. lhen when on{1 fJnwunters Ihe dial,-
noslic footprint in Ihe fulure lhe infmenee of Ihe

prior presence of Ihe indicated animal may be con-
stdered an inference of high probability.
For an inference about Ihe past to be of high prob-
ability, an eddtttonal proposition must be mel-tha1
the same releonshlps obtatned in the past as nb-
tetned in the present between bears and their foot-
prtnts! Here we introduce Ihe Interestlng and irnpor-
tant. perhaps crucial, problem erchaoologists must
solve-chow do we justify il unifnrmitarian assu!!!.l!-
,!;ionl This tssue is perhaps well illustrated through
a dlscussion of the treatment given the problem by
the pioneers of hisloricaJ geology:
Lyell's concept of uniformity h6S four majar, and ver}'
(tI NrJturijffllW!f ate:-ifurllililJ,1 (uoiform) in space-at'ftf
time. As Joho Sluart Mili Ihis jsnol o statement
aboul fhe world; jI is 00 a priori cJoim of metho<! Ihat
scenlisls mus! make jn order lo pro{:eed wilh ony
(lnolysis of Ihe pos!. lf the pasl is co.pricious, if Gad
vlo[stesn8tur811aw at will, then sciencecaooal unrllvel
hislory --
(2) now operaling to mould Ihe Barth's sur-
faceshould be invoked to explain the events of Ihe pasl
(unlformilyof process through lime), itlypres-errt'Ptb.
cellses'can Therefore, weore beHer
off ifwe con exp/ojn pas1 eveots os o result of
stil! oc/j"g. This agojn is no! on argument (lboul the
world. il ISo slofement aboul sdeollfic procedure IS l.
Gould 1977:150: emphllsismine[.
As was poinled out by Gould, the remaining two
senses in which Lyell used the concep! of uniformity
were in fact assumptions about the world, existential
in characler. One has been lergely suslained by re-
search (Le., geologic change was largely uniform in
rale, slow, gradual and sleady, not catadysmic). The
olher c1aim was a1so exislential, namely Ihal lhe
eorlh has s uniforro coofiguralion, or it has been
fundamentally Ihe same since ils formation. Mast
would agree thal Ihis has been demonslrated lo be
quite false as 8 general descriptive slatement.
What is indicated here is thal we musl make uni-
fonnilarian assumplions if we are lo gain any under-
standing of Ihe pas!. 00 Ihe other hand, when we do
so we are making empirical daims aboul Ihe post
and Ihese mUst be warranled; Ihey must be subjected
to evaluaHon. The degree lo whir.h such uni-
formitarian assumplions are warranlcd is a measure
of the degree lo which our lnferences drawn from
knowledge of tbe contemporary world andlor uUT
understandng of tts processes in the form o theortes
and laws are relevan! to the pasto
lnsofar as our tnferences regarding the pasl reer
lo the dynamics of the past, these tnferences must be
accomplsbed by appeals to principies or knowledge
about dynamics and how static propertes preserved
in the archaeological record may be derived from
dynamics. Since the only access a researcher has to
dynamics ls through conlemporary expenence. all
research directed toward the develcprnent of princi-
pies thal serve lo make possble inferences aboul the
pasl must be conducled with documenled dynamic
siluations generally in the presen!. Such knowledge
of"conneclions" between slatics and dynamics must
derive from experimental research conducted wilh
documenled living syslems.
Since knowledge of dynamics derives from ex-
perience wilh living systems, observations of linkage
between statics and dynamics musl be made on liv-
ing systems. In arder to use these principies of link-
age for msking slslemenlS about lhe past, we must
make 9 uniformilarian assumption with respecl lo
the properties used in inference. In short. we musl
assume thal knowledge gained fram actualistic
studies is relevanl and applicable to Ihe living sys-
lems o the pasl. This basic proposition must be true
ir inferences employing principies gained Ihrough
Ihe study of eontemporary dyosmics are lo be used
in inferring the past from palterned statcs. This
means lhallhe assumption is alwayscondilional and
may be false; thal is. we could be wrong in our judg-
ments regarding Ihe condition shared by systems or
entities of Ihe pasl and lhe present.
For instanee. any number of "correlates" between
sts\ics and dynamics mighl be observed in Ihe
modern world. However, the firsl question we musl
ask is whether .we are gbseryjng ao jncidence p
cause and effect, or whelher there is simply correla-
tion or coincideoce. The second, and equally impor-
lanl, question lo be considered is
posed causalion was also characlerjslic of Ihe pas!.
Both queslions must be answered affirmatively be-
fore eb8ePIatioomay",eehst-imllyserve
8& a,prfilQistlJof inl1nmces-reR8f.(Hng4he pUlI.
Although I basically agree with much of Schif-
fer's (1972. 1976. 1978) general discussion of ar-

chaeology and Ihe need for understandtng of forma-
ti0I!_processes, I generelly dtsagree with almostaf[;f
'li'i'Ssggesrros as lo how lo solve arcbeeologtcal
problema. He faits lo make the critical dtsttncuon
belwaen descr-ptlon and explanatton. This is clear as
he cites Nagel for "experimental laws" .e.. empiri-
cal generalzetons and Hempel for "covenng laws"
(i.e.. theoretlcellaws). as f these were Ihe seme thing
[see Schiffer 1976:4). Schiffer alsn argues Ihal "the
subject matler nf archaeology ts Ihe relatlonship be-
tween human behavior and material culture in all
times and placee {p. 41." 1might agree thatthls is one
wey of viewing the concerns of middle-range re-
search, bul find it hard lo accepl as Ihe cenlral focus
for archaeology since Ihe archaeological record con-
lains no direct informalion on this subjecl what-
How do we know whal experiences wilh living
systems are relevant lo the pasl? This question is
parlicularly germane wilh regard lo central issues
such as identification. Idenlificalion, as mentioned
earlier, is a key issue in archaeology, since it is this
"acl" that eSlablishes Ihe language for discussing Ihe
pasl, and in turn Ihe language carries meanings and
provides the units for logical analysis. ldenlifying
things (see Whilehead 1967:144) becomes Ihe acl of
translaling from the domain of malter inlo Ihe do-
main of ideas. It is the identities that bridge the gap
hetween lhe past and the presen!. that provide, as
Whitehead (1967:159) wOllld say, the "elernal ab-
jects," Ihe "durables," which serve as the basis for
recognizing events. fhe basis for analyzing events
and recogni2ing transitions from one evenf lo
Whalever pllsses is an even\. Bul we fnd entities in
nature which do nol ptIss. raclors in nature which
are without passage will be called obiects... recogni-
Ilon is reflecled in the intelleet as comparison. but JI
is nol Ihe ewnls which aTe compared. Jo'or each p.v"nl is
eSIlBnliaJly unique and inwmparable. What are cum-
pared are the ohjectsand relallons of obieclssiluated in
events IWhitehead 1957:124-12sl.
It seeros to me Ihat uniformitarian assumptions
function much ke in!ellectual anchors, for Ihey
provide the "points of knowledge" from whir.h we
may judge tbe extent of our ignorance regardtng
propertes of the archaeologtcal record,
Whal are the durable uncbangtng charectertsttcs
that the events of the presenl shate wilh the past? As
1 ndicated alsewhere [L. R. Binford 1977a:81.
We mey reasonably ask . whether or not thera are
classes of data remaining from the pasl wnch might
bcttcr suppcrt uniformitatian assurnptons. In short. are
there nol clesses of phenornene avatlahle tu us Ior which
a more reltebte sel uf conditiuns might be projected intn
the pest Ihan for human behavior per se?1
J answered the rbetorcal questton by suggesting
Ihal lhe sludy of the spalial slructure or the arrange-
ment of "objecls," in the Whitehead (Ig57:124j sense
of the world, would be a useful area for develop-
menl. I continue lo be of this opinion. On Ihe olher
hand, I had suggested Ihal ecological and analomicel
characleristics of Ihe species sliU extant with whkh
ancient man inlerac1ed were enduring objects for
which uniformitarian assumplons mighl be se-
curely warranted. It is hoped Ihat olhers will elabo-
mte this lisl of domains and pursue middle-range
research along as man}' diverse Jines as we may be
ab!e lo justify uniformitarian assumptions.
1 began the discussion in Chapler 1 with a clem-
onstratioll lhal !he "inlerprelalion" of cerlain aro
chaeological observalions was dependenl on a basic
premise, an assumplon aboul Ihe conditions in Ihe
past surrounding Ihe formalion of lhe deposit wilhin
which archaeological remains were recovered. I
showed Ihat Ihe assumplion was generally made Ihal
man was the agenl responsible for the disposition of
aUmalerials found in associalion with demonslrable
arlifac1s. AH Ihe "interpretations"-lhe postulation
of bear culls. cannibalism among early hominid
populalions, mass killing of elephanls at Torralba,
syslemalic hunting of hyena by Neanderlhalers at
Pin Hole shelter-were inferenlial arguments consis-
lRichardGould has argued Ihat. I;incesorne characll!Tis-
tics appear unlikely to bridge Ihe present aod the pasl. we
should avoid unifurmitarianassumpliolls "The lessIhe ar-
chaeologist must dllpend upan uniformitarian assumptions
lo iofllr past human behavior. Ihe more valid his expla-
nations will be IR. A. Gould 1978:2551," This is nonsense,
in my view, sinee any inferpnee. even a simple identifica-
tion. lo the pas! mus' make a uniformilarian assllffiplion.

tent with the Initial assumpttons. the premises upon
which the tnferences rested. There is en importan!
characterislic of al! mferantial argumenta, simply
that we can never reoson in o volid monner (rom
premrscs '0 a conclusion thcr contradicfs !he prem-
tses wuh which we stort. Ths fact has importan irn-
plicenons for archaeologfsts:
1. AII our statements about the past are in-
ferences relatve to observalions made on the
contemporary archaeologlcal record.
2. The accurecy of our Inferential construcons
uf the past is dtrectly dependen on the ac-
curacy of Ihe assumptions or premises serving
as Ihe basis of our inferential argumenls.
rhe conclusion we musl draw is Ihat we cannot use
eilher Ihe archaeological record or Ihe nferred pasl
lo test our premises or assumplions. Quite Iilerally.
al! our reasoning is "Iocked in" by our original
premises and observationallanguage. lJnless we can
lake our premises lo experience and permil experi-
ence lo pass judgmenl on Iheir accuracy, we can
never gain a critical perspective wilh regard lo out
beJiefs aboul Ihe pas!. "Can we presenl hislorical
events as instam:es ur confirmation for a law? We
cannol if Ihe very law we wish lo tesl has becn pre-
supposed in inferrillg the event IKitls 1977:791."
Pul anolher way. since we construcllhe past in-
ferenfally ..... e cannol use our conslruclions lo test
Ihe accuracy of Ihe premises Ihat provided the basis
for Ihe chara(:lerislics conslrucled
Since we cannol use lhe inferred characlerislics
of Ihe pasllo lesllh!'! basis for our inferenlial proce-
dures, ho..... do we develop reliable meons fur know-
ing lhe past? The answer, as 1have intimaleel, is Ihal
wp, must engaBe in middle-range which
consisls of aclllalislic sludies designed to conlrol for
the relationship belween uynamic properties of the
pasl abolll which one seeks knowledge and lhe slalic
material properlies comrnon to lhe pasl and Ihe pre-
sent. Whileheau's "eternal objecls"-in shor!. lhe
charaeleristics about which uniformitarian assump-
lions may be made, Ihose Ihings which lhe presenl
shares wilh lhe pas!. These cornmoll Ihings provide
the hasis for a Lornparison of Ihe events of Ihe pm-
senl wilh Ihe cvelll:- 01 Ihe pasl or evcnls from (Hf-
ferenl limes in I!w ]las!.
The reason that middle-range research must be
bastcally aclualislic s thet only in the present can
we observe Ihe bear and the footprtnt together. the
coincidence of the dynamic and the static deriva-
ttves. In more mature disciplines, where a releuvely
sound melhodology and a eophstcated observa-
ttonel language exist, it may be possfble lo use in-
ferred conditions ebout the pest as premlses for fur-
ther tnferences if the inilial premisas servng as Ihe
bass of the original inference are securely
documented and "vertfied" at the mddle-range
level of research. As illustrated in Chepter 1. this is
probably a very rtsky strategy. given the lack of
sophislication in conlemporary archaeology.
The dependence of our knowledge of Ihe pasl on
inference ralher than direcl observalion renders the
relationship between paradigm lthe conceptual 1001
of description) and Iheory (lhe conceptual tool of
explanalion) vague. it a1so renders lhe "indepen-
dence" of observations from explanalions frequenlly
suspecl and comrnonly standing in a built in rela-
lionship, Ihereby commilting Ihe fallacy of "con-
firming Ihe consequenl."
It is this condition Ihal renders it imperative that
our methods for constructing lhe p8!sl be- in-
lellectllally ndependenl of our lheorias for explain-
ing Ihe pasl That is, Ihe lheories explaining the ar-
chaeological record. lhe work Ihal provides our ab-
servational language and conveys meaning lo ar-
chaeological phenomena, musl be inlellectually in-
rlependent of our a priori ideas of lhe past. or our
lheories regarding lhe processes responsible for pasl
evenls, pallerns of change, or slabilily.
range .fheoq"mt.lsj...he inklUec-tually;
Middle-range Iheory musl be
lesled primarily with documenled living systems.
Middln-range lheory Ireals Ihe relationship belween
slalics and dynamics, between behavior and malerial
derivalives. General theory may be lested using ar-
chaeological phenomena meaningfully oper
ationalized Ihrough middle-range research, Stated
anolher wa.,'. Iheory musl be evalualed
instrumenls for rneasuring lhe variables specified in
Ihe Ihenry, Thcse inslrumenls musl have been de-
veloped indepencienlly Ihrough middle-range re-
seflrch In lhe absence of melhods for reliably
l1lolliloring the variables silid lo be clelerminantly
, !I
11, ,
1: .;

2. Middle-Hange Research und the Role af Actuolislic srudes

~ ,
' 1 : ~ 1
I . '!'(
'1 i!
I 1:1
" J
'Ii, ,
1" ,
Part II
search of
In Chopter 1 1argued that a par-
ticular set o{ assumptions cbcut the
jormctlon 01depostts that contctned
archaeological remoins served as the
warrant for further inferences re-
gurding mctericls found in nssoctc-
tion with the orchoeological re-
1 attempted lo shcw that
Paleolithic archaeology developed
in a sttuouon where a specific con-
cern [or methodclogtcol resenrr-h
was no! seen as sepcrote from re-
search conducted for purposes al
learning about 'he post. Interprete-
tions were lorgely developed post
hoc or after dscovertes had been
made in the orchoeclogiccl record.
These procedures Iorgely ccnslsted
a[ using inferences based on os-
sumptions regording the formation
processes cperntlve in rhe post or
the condHions responsible [or mor-
phoJogicol properties or potterns
of associotion observen in rhe nrchne-
oJagico/ record. Once snch posl hoc
not evalate our ideas ebout the pest and why it was
the way it appears lo have been wtthout means of
montorng Ihe eonditions or variables belteved lo be
important. Both of these tasks are dependent upon
the development of mddle-range research.
opereuve. no archaeologtcel test of general theory is
"rne conc!usion should be clear: Middle-range re-
seerch. wlth particular emnhasts of theory building,
is crucial lo the Iurther development of arcbaeology.
We canno! "know" the past without it, end we can-

'1'1 I ,


11:1: f
~ I ~
'1 ~ i
Chapter 3
Patterns of bone
produced by
nonhuman agents
The observettons surnmarized in thrs chapter are
largely concemed with Ihe modficatfons that ani-
mala. considered strictly as rnechancal agente. make
on bones. That ts. 111m concemed with looth mllrks
and patterns of differential destructon that gnawing
animal s are capable of producng and in facl regu-
lady produce. In all my descrtpuons l am concerned
wh recogntang modificalions of bone that are dis-
tinctive or otherwtse dtegnosttc of action by nonhu-
roan agente. I cannol emphasize too mueh that with
anirnals most bone modificalion s etfected by teeth.
Teeth are the rnplements and the aw is a mecha ni-
cal device Ior exerttng force; Ihe Iwo constitute a
dlstinctive mechanical system. This s consdered lo
be a very non human mechancal situeton, in thal
man commonly employs a varety of Irnplements.
such as kllives and hammers. end the mechaniral
situaticn is one of dtrected culting or tmpacted
hammering. For entrnels. the teeth are organizad in
whal may be vewed mechancally ee a vise. These
are real differences in the mechanical besrs for
human versus nonhuman predator-scevenge-
modtcaton on bones. Most of my descriptions
focus on recurren! propeettes beleved lo refiecl
Ihese mecbentcal differences.
Obviously animals differ in stae. feeding hebns.
the mechanical strength of their jaws. and the destgn
of thefr teeth. Nevertheless. they 811 may be constd-
ered animaled vises.1t is expected Ihfl! the baste pat.
Poli JI. Middln-RunlW R{'.<f'unh- -In S/lon:h nf McthodoJogy
tntcgroto theso observolions with those of others regarding the behavior af
predctor-scovengers. These discussfcns will then provtde the back-
ground jor o consideralion of csscmblcge variabilily, thct s. the determt-
nants of patterns of ossociuuon omong difjerenl anolomicaJ par!s in
terms of animal bebcvlor contrusted with thct of OWIl. From thls 1 wil/
drow sorne inlellec!uol expectctions cbout the behavior of predator-
sccvengers. "cnchors" thut we can project into the posl and use fa iden-
tijy the conditions Ihu! generaled cerlain conjfgurottons offaunaJ mate-
ria/ remclrung for us lo observe todov.
""/(YJ. -
&' ......-

3. nJ Bnne ModiJiroljons Prlllilll"ed by Nonhumon

I have pictured olher exomples of bontl lnlerpenetro_
tOIlS ill Ihe memolr on Osleodonlokeratic Culture
such as on alllelope co!can;,us wedged between the
Zygomotic arch and the lateral aspect of a split hyoenB
skull. and flekesofbone or whole boneswedgBd withill
other bones. These (l)l;amples are here shown lo be lypi.
cal and are nowsufficiently numerous lo conslitute ir.
refutable cvldence of a systemotic auslralopithecine
technique not of smashing up skulls bul also 01
split1ing by m(lans of driving olher flakes of bone or
whole bones forcibly down lhe shahs of olher bones
IlJarl HIfiO:1361.
New World erchaeology in Ibis survey beceuse iI
provides an nteresttng case of research mnving 10-
werd the development of a methodology that. as I
will attempl to demonstreta, is largely poorly
founded and mtsleedrng. II has the added interest
that many of the properties cted by New World ar-
chaeologists have been clted by Africanists and other
Old World archaeologists as diagnostic of human
behevor. The very fact tbet many of the seme formal
properttes hava been observed in widely separaled
places atld with reference lo differitlg lime periods
adds to their middle-range research potentia!. If
these properties can be underslood accurately, then
Ihey are good candidates for strong methodologkal
tools, since making secure uniformitarian assump-
tlons about their meaning is encouraged by Iheir
broad spatial and distribulions.
Most claims for man's use and modification of
bones as tools have been based on cited evidcnce of
either altered shape or particular modifications such
as "flaking," "retouch," "polish," and "abrasion,"
which were believed to be producls ofhuman action.
In addilion there were e1aims made Ihat man broke
bones differently from animals. res\llting in distinc-
tive and diagnostic patteros observable among Ihe
fragmenls of long bone. As was poinled out earlier.
Breul {1938:581thought thal roan broke bones "Ion_
gituditlally" whereas animals broke bones trans-
versely. This view was accepted by Weidenreich
(1941: 74-77) and ciled as evidence that Sinan-
thropus was a cannibal. Raymond thIrt (1960) ac-
cepted Ihe suggestion of 10ngitudinaHy fraetured
bone splinters, and proposed thal this Iype of break.
age was accomplished by wedging bone 1001s ioto
exposed bone cylinders. thereby splitting the shafts
of long bones longiludinally! (for criticism see Feus-
lel 1969. 1970.)
PrevtousApprocches lo Underslandinglhl' SigniJicance of Broken ond Modificd Bone
The sample of dog-modified bones te considerad
unambiguously informative ebout dng behavor wilh
raspect to Ihe following properties: (a] surctal
modcetons and lb) pattams of deslruction or
breakage of single borre unfts. Prcperties not consid-
ered referable lo dog behavor are (al the relative
frequences of the bones present and, lb) differentlal
destruction due to Ihe accessbtltty of the different
bones-Ihal s. dtsaruculetton hed alreadv been ac-
comphsbed bv rnen befare Ihe dogs gained access lo
tha anatomcal pars. so patterns of deslruclion were
nol generally condilioned by the patterns of anatom-
ical arrangemenl characleristic of a prey species.
Thus the population properties of animal-obtained
assemblages cannol be undersloorllhrough the sludy
of this materia!. On Ihe other hand. variabilily in
moc1ification and destruction of given analomical
parts is documented in a reliable manner.
Again. I am confident that the properties de-
scribed are referable to the Betion of dogs. The beBr
and the footprint are together wilh regard lo
modifications of single bane units. However, 1can-
not USe these data for studying how dogs would
modify larger unHs of analomical organizalion, such
or a whole animaL
:... AH offerad are wilh regard toJ
caribou bone. I will iIlustrate sheep bone occasion-
ally lo indicate Ihal Ihe generalizalions about
caribou also ap'ply to sheep. Statcd another way, I am
talking about the modification of caribou bone by
Eskimo dogs}largely malemiuts-weighing be-
tween 55 and 70 lb. I will frequently cite lileralure in
whch olher researchers have described similar
modifications of either bones from other spedes
[horses. bison. etc.) or banes modified by olher
Bgents (large cats. hyenas, coyoles, etc.). These refer-
ences serve lo warrant the opinion thal the
modifications produced by dogs may be generalizad
lo other species of predator-scavengers and to the
bones of olher prey that might be attacked by gnaw_
ing animals.
Previous Approaches lo Underslanding Ihe
Significance of Broken and Modified Bone
As was shown in Chapler 1, I seriously question
many of the daims old World archaeologisls pre-
senled \he pas!. However. I concpnlrale on
bage" or tabie scraps. but whcn this general peuem
is vlnlated it s never with raspee to prcvouslv bro-
kan long bones. When bones that have been pre-
viously "consumed" by humana are gtven lo the
doge. they are almos alwavs "women's bones" (see
L. R. Binford 1978b:149) or those that have been
boiled. These nclude the neck. Ihoradc vertebree.
rbs. lumbar vertebree, pelvis. aud sr.apu!a. 'I'he only
modificalions that might occur lo these bones prior
lo roodcauon by gnawing dogs would be cut marks
rorn ects of dtsmemberment. chop marks for sizing
the bones to nI in a stew poI. and sometimes
"mashed edB!;s" resultill;; from human chewin;; al
the margms of lhe bones during consumption. Feed-
iog dogs womeo's bunes is much more common 00
winter sites where the warm bones are viewed as a
treal for the dogs. 1 never observed Ihis table-scrap
feeding in rny spring or summer dog-feeding rec-
ords. Generally dogs are fed complp.tublllr.hering un
its, Ihal is. a complete rib slab. a complete scapula
wilh al1ached meat, a complete pelvis, aod so on. In
butchering the Nunamiut dismember the body of a
prey animal by cutting between the articulator sur-
faces of Ihe bones. Somelimes multiple bone units
will be relurned from the field lo the village. These
may be further butc;hered al !he time of dog feeding,
bul except in the winler, when the meat is frozen.
secondary butchering is the same as primary: dis-
membermenl by cutting between the bones al the
joinls. This means Ihal ;ingleor mulliple bone unils
are whal Ihe dogs receive as food. Thev are almosl
never presented wilh previously broken bones. The
exception lo Ihis generalization is feeding during
winter when meal stored outside is always frozen.
Secondary butchering under such conditions is ac-
cornplished wilh an ax and it s common to break
lhrough the shafts of long bones as a regular dis-
membermenl strategy. For Ihese Teasons. three of the
dog yard samples (see L. R. Binford and J. 8. Bertram
1977:80J listed (samples 1I, 1Il. and IV) in an earlier
reporl were eliminated from this analysis Rnd only
consulted concerning patlerns of deslruclion of ar-
liculator ends.
It should he pointed out that 1have analyzed over
122,000 bones known lo be Ihe result of Nunamiut
behavior. I have nol reported any type of modifica-
lion as referable to the dogs Iha! regularly occurred
in the known "human sample" (!\ee L. R_ Binford
terns of modcatton thet en anirnal is mecbantcallv
careble of producing are nol gong te vary 8 great
deal from rme species to enother. lt ts more likely
that the magnitude of the capabtlty. and the use of
t. will varv emong spectes. A very strong animal
may be 10 accomplish the destrucnon of large
bones but the tecttc o destruetion s expected lo be
essentially hke that of a smeller animal addresstng a
smaller bone. This ls an operating assumption guid-
ing the presentaton of much of the following mate-
riaLI will remain sensilive lo Ihe possibility lhal lhe
assumption is overly simplistir:.
The demand thal the bellf and ils footprinl be ob-
served together is only parlially mel for 'he IDa\erials
described. o de-
scriptions and iIlulItrations of modifications pro-
duced by animals WBre recovered from nine dog
yards within village of Anaktuvuk, Alaska. A
total of 12,716 bones were observed for patterns of
surficial scarring, bone breakagc. and destruetion.
An additional 416 recovered from two small wolf
dens served primarily to permit lhe assessment of
the impression Ihat Ihe pat1erns produced by the
domestic dogs were not unique 10 the dogs.
The samples from the Anaktuvuk dog yards wefe
not collected wilh Ihis particular research aim in
mind. The original mo\ivalion was to study in detail
Ihe seleclion of parts by Ihe Eskimo for use as dOR
food and to monitor any seasonal changes in the
feeding strategies (see L. R. Binford and J. B. Bertram
1977: 78-g01. These dog yard samples werc supple-
menlal lo actual records of dog feeding and follow-
up studies of the bones remaining al feeding spols
after the dogs had been moved away Isee 1. R. Bin-
ford 1978b:2B2-265). In addilion, 1 had conducled
extensive interviews with Ihe families then keeping
dogs regarding their dog-feeding lactics.
Thus Ihe sample has a weakness. Man actually
inlroduced lhe bones to Ihe dogs after he had killed
the animal and transportad Ihe parts to Anaktuvuk.
Thal is, aU the characteristics of Ihe bones canool be
unequivocally refeITfld exelusively lo Ihe behavior of
the dogs. Nevertheless, Ihe characteristks described
can be ascribed lo Ihe dogs with Ihe highesl levels of
confidence. For inslance, my informal ion no pal-
teros of dog feeding by Ihe Eskimo perndls me to say
with confidence thal no bones previously broken for
marrow within Ihe viHage were given lo the dogs.
The Eskimo generally do nu! fpR.d their dogs "gar-
+- .....-f
l-W1A"j. -

-y .,
, .
J. Potlerns of Bcne ModificationsProduced by Nonhumcn Agtmts PreviousApnrouches lo Underslat1ding the Significanceof Brokenand Modified Bone
The femous crack-and-twisl technique of bone
fracture was proposed by Dart nt as a method of
fracturing bones for maITOW extractlon. but instead
as a meens of tool production, 8 postton leter
adopted by Saclek-Kooros (1972).
we were both gnorent ot the means wbereby primitiva
saplent man at Kalbank 15,000 years ago end IhE! sUll
more primitiva protoman AustroJopithecus prcrnetheus
ebout 1,000,000 years ago al Makapansgat obtained
theee spral blades unti! one Sunday we preservad the
femur Irorn Ine family's roest leg of mutton.
Th8 following doy he lKitching] brought il lo the
laborl'llory, slruck the shaft with the poio! of a stone
implement and holding 80ch end of the bone in a hondo
twisted Ihemopar! in opposihl directions. Tha beautiful
resultanl spiral blad611 and Ihe poinl o impact of Ihe
stone implement are clearly see".... So the nexl time
this Sunday oinl provided ,ames Kilehing wilh a fur-
Iher experimental opportunily he simply struck the
shafl against Ihe wooden edge of the lable and Iwisted
the bone ends contrary wise again and the spiral and
pointed results ttnl seen {Dart 1959b:911
points.s:ho,,!ld_be clY!!fflJ!..IDLm. Exper-
iments have repeatedly demonstrated (L. R. Binford
1978b:152-157; Bonnichsen 1973:14, 1979:37-53)
that plltl f etttre resull. from breakins rel8tively
mm )yuw:1t is not conditioned by twisling. Twist-
ing may condition two types of situahon; most likely
it simply facilitates the separalon of fractured bone
from the periosteum and in no way condlions the
pattem of actual breakage. In siluations such as are
described by Sedek-Kooros (1972) where a line of
minor impacts are made longit\.ldinally along abone,
the twisting may actually ensure a longitudinal
crack connecting all the small poinls of impacto This
is difficult to accomplish and mosl often twisting
simply removes or separates fractures that are oth-
erwise held togelher by the periosteum. (See. for in
stance, such fractures indicated in experimenis 19
and 22 IFigure 51 of Sadek-Kooros 11972:3741.) De-
spite Dart's designalion of Ihe crack-and-Iwisl
method as one used in tool produclion. il has been
most often cited as a melhod appropriate to the
breaking of bone for marrow exlraclion. For instance,
as a caption to a very imaginative reconslrucHon of
the Iifeways of AustraJopithicus ofriconus in the
popular Life Nalure Library book, E"orly Man, by F.
Clark Howell (1965), we are lold,
Cracking and twsting antelope limb bcnes. Ihe Au-
slralupilhccines estract the edble lnsides. Late the
bones. frequently found with prehuman Iosstls. may
haveserved a different purpose: their sharp erigesmade
useful tools and weapons [Hnwell 1965:69).
Recent New World Htereture. atthougb generally
acknowledging that twisling s not necessarv lo the
producton of spiral fractures. has tended lo suggesl
thal spiral fractures themselves are dtegnosno of
human behavor. This view seems to slem from sem-
inal work by Bonnichsen (1973), which builds on
two ploneer reseerch experlences with behsviorally
modified bones. Bormichsen participaled in a sludy
with Zierhut (1967) aimed al Ihe elhnoarchaeologi-
cal study of bone breakage 8mong Ihe caUlrrg
Cree. This experience was coupled with the analysis
of bones recovered from animal cages al Ihe Alberla
Game Fano in an importanl alulIupi todevelop
criteria for dislinguishing between bonebroken or
modified by man and that broken or modi-fied by
animals. Research with Ihe Cree involved repealed
demonslrations by the locallndans as to how bones
were broken for marrow (see Bonnichsen t973:Plale
1). A standard melhod was observed 8mong the Cree
and Bonnichsen obviously still accepts this Isee
Bonnichsen 1979:Plate IV-1) as lhe method wherebv
man break bones for marrow. .
Bonnichsen's second actualislic experience was
in lerms of observations made on bones recovered
from Ihe ca;;es of a caracal cat, a North China
leopard. a Siberian Iynx, and a Siberian tiger (Bon-
nichsen 1973:14). This lisl is interesling in thal it
ineludes no can id or hyena. Mosl observen of cals
agree that Ihey are fastidious ealers and modify
bones primarily in the course of removing meal from
Ihe skeleton rather than during the gnawing sessions
more characterislic of canids and hyenas. Large cats
have, neverlheless, Iremendous power in Iheir jaws
and are capable of shearing bones during Ihe course
of meal consumption. (Sea Ihe impressive pholo-
graph in Beard 11979:961.1 AH of Ihe bones ilIus-
Irated by Bonnicbsen (1973) and attribulable lo ani-'
mals at Ihe game farm werc modified by a Siberian
tiger, cerlainly a powerful animal. AII of the
modifications iJIuslraled are puneture marks on
bones of a small animal (sheep). whir-h when broken
were simply sheared by Ihe Hger (Bonnichsen 1973:
Plate eb. c, d, and f). In one sheep libia illustratecl,
however. Ihe proximal end was sheared off ancl a

spiral fracture originating at the end of the bone was

The last "control" population of bones availeble
lo Bonnichsen was a colleclion recovered rom the
stte inhabited by the Calling Lake Cree, who had
servad as informants for the marrow-cracking dem-
onstrettons. These bones represent gerbage. from a
local house. Ihal was subsequently scavengnd by
domestc dogs. Thus, iI was not cleer which egent.
man or dog, produced the modiftcetions within ths
population. Since man obviously crackad bones for
marrow, the spral fracluring within Ihis assemblage
was attribuled to man. Only modifications pre-
viously described by SulcIiffe (1970) and Miller
{1969} as characteristic of animal modificalion were
illuslrated as referable lo Ihe dogs. Thu5 the
"modero sample" from Ihe Cree was interpreted
using previousiy described criteria for recognizing
animaJ-modified bone, coupled with Ihe elhno-
graphic experience of marrow cracking among Ihe
Cree. which was token as the basis for recognizing
bone broken by mano Dnly a single piece of "am-
biguous" evidence existed from Bonnichsen's
perspectiva: the spirall}' fractured sheep tibia from
the Siberian tiger's cage,
Sornebanes alterad by animals axhibit spiral fraeturing
such as the sheep tibia from the Siberian tiger
cage.... Asthis process has nol beenwilnessed,the fol-
lowingsuggestion is conjeelural. The bone WIlS proba-
blyHrst perforated by the 'eelh at the two oppositional
points11 the epiphyseal end. Al the same time lhe liger
pulled out a segmenl whieh terminates in a spirlll hllc,
lure IBonnkhsen 1973:t61.
Bonnichsen then concludes that tbe long bones are
particularly important for recognizing human activ-
ity, noling Ihal there is frequently an impaet notch
and spiral fractures initialed outward from Ihe
notehed loading poinl. In con Iras!. bLlisl0.!!!!..fu!..
10wJ!!&... as of animal-
modified bone: "(1) toolh perforalion marks. (2)
gnawing and scooping oul of cancellous lissue, (3)
crunching and splintering, (4) spiral fracluring di-
recled from Ihe epiphyseal ends. (5) partial digestion
[Bonnichsen 1973:241."1
IHow Ihe Ihird chllrllcteristic is different from spirlll
fracturing is not lit all dear since this criteria is adopted
from Sutdiffe (19U) and hE' did nol ulilize lhe concept or
Bonntchsen was led lo believe that spiral frac-
tures were characteristc of bones modified by men,
whereas tooth marks. scooptng out. and parfial di-
gestion were charaetenstic nf animal.modified
bones. This is imporlanl, stnce Bonnichsen obvi-
ously beleves that man breaks marrow bones by
smashing them in midshaft. the "mid-diaphysis
smash technique of spiral fracture lBonnchsen
1979:36[." Althougb be admits the posstblny hat
animals rnight produce spiral fractures originating al
the epiphyseal ends. it is quite obvious thal Bon-
nichsen constders sptrel fracture lo be "artificial."
He slales,
The raeognition of spirlllly fractured bones has led to
the following hypothesis: splrlllly fractured bones are
indicative of man's presence and renect patlerned
human behavior. This hvpothesis has been used lo
guide Ihe analysis of faunal remainsfromthe
Old Crow Flats palBOnlological assemblages. Two
additional hypotheses specify how spirally fractured
oonescan be Usedforreoonstruclingprehistoric human
behavior pafterns. Hypothesis 1 states' The
dislributions of spirally fraclurad bones may indicale
human aclivity localities;and, Hypothesls2 Slales: Spi-
rally fractutad mammlll bones ndicate the general
seleclivily exploited by man [Bonnichsen1979:69].
Methodologically speaking. thougb Bonnichsen
refers lo his propositions as hypotheses there is no
way of lesting Ihem using paleontological remains.
since Ihey refer to a relationship belween a property
remaining today (spirally fractured bone) and an
agent inferrad to have been aclive in lhe remote past
fman). Since he is attempting to use Ihe spirally frac-
tured bone to prove the ancient presence of man in
the New World. only with an accurate and indepen-
denl means of monitoring tba ancient presence of
man could he tesl his hypotheses. In fact bis hypolh-
eses are used as conventions for giving meaning lo
paloontological remains. They serve to ustify his be-
Iief that man was presen! in very early times bul in
no way do tbey evaiuate tbe belief. That Bonnicbsen
does nol seriously consider predator-scavengers as
agenls capable of producing spiral fractures is iIIus-
lraled by his discussion of possibJe sources of spi-
rally fractured bone:
The prehistorieoccupants of OldCrowFlatsapparently
processed the limb bones of large mammals for "bone
illl I
., '1
'" I
:i1 ,
'11' :
! ji

3, Pcnerns 01BOIlf' Modifil;:llliOnS Prcduced by Nonhumcu Agents
Previnus Approllches lo Understcndmg the Significance oi Brokenond ModUied Bone 41
I i I
, i
grease" end perhaps for tool productlon material. Five
bundred Iiy-five spiraJly fractured limb bones trom :'0
loclllilies suggnst that bone pmcessmg WIIS widely prec-
tlced throughout "the rtets.':
ln essessng spiral fractures, Ihe possibility should
no! be rulad out Iha! occasionally enimals broke ther
Iimb bones in spieal fractures . Three argumenta
suggest Iba! he majorily of spirally fractured bones
Irom Old Crow Plats are bones broken bv man. Of the
1,794 elements collectecl from loco.lily 89, 237 speci-
mens exhbt aptral fractures .. This figure is much
higher than would be expected. if only natural breakege
were involved. Secondly. bOlles exhibiHnll spiral frac-
tUteS nelude: marnmothimastodon. horse. caribou. hi_
ron, muskox, elk and maase. Spiral fractures wr.re nol
observed on Ih6 limb elemenls of bear, symhos. saiga
antelope. camel, cal, or ground sloth ... and the third
argument is; the bones exhibiling spiral fractures Ol;:f:ur
on all the majar Jimb bones. lf !he bones were broken
only by natural causes. B paltern in which Ihe weakesl
limb bones were broken would be expetted (Bon
niensen 1979:77-791.
We see here a c1assic argument from elmination
and a good illuslration oC ils weaknesses. Two causes
have beeo postulated: (a) man breaking bones Car
marrow and possbly for tools and (b) animals aeci-
dently breaking their legs. The laiter is rejected by
arguing that accidental breakage oC legs should be
rare, and it represents at leasl 13% of lhe observed
bones at one location al ald Crow. 1 totally faH lo
undersland the logic oC Bonniehsen's second poio!,
but the final point argues Ihat aH the major lmb
bones exhibited spiral fracture though accidental
breakage would be expected 00 only the weakest
bones. Having rejecfed animals breaking their legs as
a cause, we may accept the remaining "working
hypothesis" that man was Ihe causal agent. Such
loglc only works if oDe has all Ihe possible causes
identified and included in the elimination proce-
dure. As will be shown, Bonnichsen did nol have an
importan! cause lisled-nonhominid predalor-
Al approximalely the same time Bonnichsen's
carly work was done, George Frisan (1970) was
sludying Ihe faunal material from Ihe Glenrock Buf-
falo lump. He excavaled in delail the locations yield-
ng slone butchering tools and assodated faunal re
mains, obviously the remaios of a mass kill of bison
that had been bulchered and processed for meat by
mano Frtson ruede the assumpuon. common in many
ou World studies. that the charectersucs of male-
rtals associated with unquestioned lools rnust also be
refereble to Ihe ecuon of man. Therefore, all the pal-
temed modifications on Ihe bones from Glenrock
were inlerpreled as the cnnsequences of butchenng
behevor by mano As I will ilIuslrate throughou! rny
discusstons of animal-modified bones. 1consider the
major modtceuons illustrated and descrbed by Fri-
son [197U) lo have bean the products of scavengtng
animals twolves andror coyotes). who presumably
Iaasted at Ihe site after man had taken his share from
the kills.
Near the end of Chapter 1, 1argued that il seemed
extraordinary lo me thal the interpreters of the sile of
Torralba made the implicit assumptioll thal nalure
slopped excepl when andenl men were on the scene.
I pointed oul tha! such an assumption was necessary.
since the interpreters refarred aH the faunal conlents
of Ihe Torralba deposits lo the aetions of ancien!
men! After having reviewed the types of car-
nivores can leave on faunal assemblages, il seems
equally amazing to me Ihat of alllhe mass bison kills
and megafaunal remeins excavaled by archaeologists
from the American Plains, nol one such location is
reporled to have yielded evidence of predalor-
carnivore behavior! AIl bone modifir:ations and palo
lerns of dispersion have been sYlilemalically Ireated
as s!emming from human behavior. This situation is
truly remarkable, sinee both wolves and coyoles
were common, even plentiful. on Ihe American can-
tinenl unlil relatively recenl limes:
A woUer spent eaeh day in the same muline. In the
aflernoon he would ride out and shool two or three buf-
falo and lace Ihe carcasses with strychnine. The nexl
morning he would return lo dress oul Ihe len or twenty
victims. One wolfer. Roberl Peck, left ti record of hs
days in 11ourna) . Peck mustered out of the IIrmy in
1861. and he and two friends set up a wolf Olmpabout
twentvfive miles norlh of Fort Larned. Kansas.Over tne
winfer Ihey killfild 3.00nanimals: 800 wolves IIml more
lhan 2.000coyolas.and Ilboutlml foxILopez1978: 1781
If we believe Ihe archaeologisls. we must imagine
thal the North American wolves and co}'oles
avoided the edble remeins left by North Americans
al such places af> animal lraps and the bison jumps
profusely documenled on the Plains. (For sorne idea
of the scale end distribulion of such features see
Davis and Wilson 119781.)I heve searched in ven for
discussions of bcne modifiratlon and dispersion re-
ferable to !he behavior of nonhuman scavengers al
sueh obvtoue ooncentratons of meato
The majority of the patterns of bone deslruction
ilIustrated by Frisen (19701 from the Glenrock site
may be taken as a c1inic in canid-modified bone.
Nevertbeless. I am gelting somewhal ahead of my-
self. The poinl ts that borres showing pertems of de-
structon cheractertsttc of cnida were mxed with
humallly modified bones in the assemblages avail
ab/e to Bonnichsen. and at Glenrock canid-modified
bones were misinterpreted by Frison as Ihe producl
of human actions. Thus. from Ihe perspective of de-
veloping melhodolagy, spiral fractures were viewed
by Bonnichsen as diagnostic of man and all the
modificalioos produced by canids at Glenrock were
advanced by Frison as also diagnostic of mano The
latter were interpreted as deriving from a particular
method of butchering (muscle stripping) postulaled
to aceount for many of Ihe modifications, whereas
others were viewed as ntenlional modificalion of
bones for use as "expediency tools" (Frison 1974).
The work of Bonnichsen and Frison provided Ihe
conventions for giving meaning lo observed
modifications. on bones, the principies later sum-
marizad as follows:
1. "Spirally fraclured bones are indicath:e of
man's presence and renecl pattemed human be-
havior IBonnichsen 1979:691." "Bones that exhibit
spiral fraclures from non human sources are usually
Ihose brolten afler the bone was dried or hao been
parlially decalcifed, ar borkl;!n when the animal fell
or Iwisted ils leg IStanforo 1979a:117)."
2 '"Bone expediency tools (Johnson 1977 and
Frison 1974) are manufactured by producing a spiral
fraclure anO using Ihe sharp edges of Ihe fracture for
butchering or hine-working aclivilies. Hones used as
1001s can be identified by the oceurrence of liny step
fraclures, polsh and slrialions on Ihe proximal end
of lhe spiral fradure.... Expedience !on[s are \ISU
ally made from long bones. bul ribs. mandibles,
scapula and innominate fmgments ware used
IStant'ord 1979B:10RI."
Using spiral fraclures. and R variely of surfkal
scarring and bone rnoliificlions as di;l.:nostic of
fauna! assemblages produced by the hand of man, it
ts nol surprtslng: gtven the vew that rnost of the
"diagoostic crileria" were probably produced by
gnawing entma!s. that a number of modern inyths
have begun ro emerge. Using the Bonnchsen-Frison
entena for recognizing man and tools. meny er-
chaeologists have made claims for man's great an-
Iiquity in North America (Bonnichsen 1978, 1979;
Canby 1979; Friedman 1980; Harington et cl., 1975:
lrving 1978; Irving and Hanngton 1973; Morian
1979: Slanford 1979a).
Similarly, there have been c1aims Cor 1001 as-
semblages lacking stone or other generally recogniz-
able products oC human workmanshp such Ihat al
least sorne workers have entertained (he idea Ihat
eariy man in the New World manufactured tools
primarily of bone:
Irvlng furthar suggested that a sophlsticated Upper
PaJoolithic varlely of nint industry is by no mean' es-
sential lo supporl human life in the far north. In fact,
bone forms that could have performed all the necessary
tasks of hunting, piereing, butehering. skinning. onrl
per[Qrating have besn recovered [rom Ihe deposils al
Old Crow and elsewhere, adding another dimension of
possibility lo Ihe reconslruction of oarly lifeways
IHumphrey 1979:viii. parophrasing WiIliam Irvingl
This sturly sugges!s heavy lImphasiswas placed on
bone working in MidWisconsin times. This pattern
probably dominated Ihe tool making piclure in many
oreas unti! Pleistocene extindion began lo oc-
cur.... Man probably responded by developing new
lechnologic.al innovations Iho! emphasized the use of
lilhic llrtifacls. In lghl of these findings. future early
manresaareh . might well be dil't'Cted towordreeam-
ining existinx Pleistocene psleonlological collactions
for evidenl;:e of human modificalion (Bonnichsen
In the hands of generalizing early man enlhusiasls
advocBling a particular point of view regarding the
peoplinR of the New World, the material from ald
Crow Flals is charaderized as "a remarkably sophis-
licated and specialized bone and anller induslry
IBryan 1978:3231,"
I consider most of foregong argumen!s
having produced a series of modero mvths. I will
tiempt to demonslrate Ihal lo) spira-I fracture is not
unique lo man, (b) man does nol generally break
bones by Ihe "mid-diaphysis smash lec!mique," (cl
.,4.d t: -
Jtc'. _

J. Pcttems of Bone Modifications Produced by Nonhumon Agents Skeleto/ DisarticuJalion 43
"Table adoptad fmm Hll1 1979:742. (Tabla oTiglnally tillad "Sequance of Disartlcula!lon in Oamalascus Korrigum").
'[xceptions lo skelala1ltfoupings are indicaled by parenlheses.
I had the opporlunity to observe and record the
parts remaining on 24 wolf kilI sites (reported in
Chapter 5). The overall paltern observed was genero
ally consistent with tha pattern reported by HUI. In
several cases there was elear indication that the
TABLE 3.01
Sequence of Natural Dismembennent AccOrdinll fo HilJ"
sequence of disarticulation. For instance, when the
forelimb separales, il is most Iikely Ihal all articu
lated segments of Ihe fronl leg will be dragged off.
not jUst sorne desirable or high-yielded elemenl of
Ihe fronl lego
II : i
Ili l'
3 2
6A 1A and B
13 6C
8 a
10B 15
l2 It
14 ItB
Next occurrence
nemed unil
BasicaIly appendicular skelelon
Unil (smaller one observed
(Craniumseparates fromatlas)
(Tarsals) separata from metalarsals
Ribs (average value)
Thoracic vertllbrae(13) + lumbar + innomnate + sacrum
Thoracic vertebme (2) lo 112)
Thoracic vertebrae (13)
Lumbar vartebrae (1) + sacrum
Lumbar verlebrae (1) lo (6)
Sacrum separates from lumbar
Thoracic vertebrae (1)
Cervical vertebme {3) lo (7)
BasicaIly axial skeleton
BasicalJyfront legs
1 Forelimb(induding scapula)
2 (Caudal vertebras)"
3 Scapula
4 (Mandible)
5 Humerus + radius + cubitus
6 (Cranlum + alias)
7A Carpals
7B Metaearpal
IIA Humerus
68 Distal phalangea or forelimb
Bosirollyreor Iegs
Be Hindlimb
9 [proximal phalanges and distal phalanges [frontl)
loA (Redtue and ulne] seprate
lOB Femur + tibia
l1A Tarsals + metetarsals
11B Phalanges (nterphalangeal joints as well)
12 femur separates from tibia
The eequence [gtven in Table 3.01)Is descrbed in terms
of Ihe smaller untt. often B single bone. resulttng Irom
the dislrti!::ullUon of a particular otnt. Becb diserttcule-
non has been gjven B number. shown on Ibe let. ln-
dil:ating ita stege in the sequence. Apperentlv simulo
teneous disarticulalions ... ere distinguished bysupple-
mentary letters. The numbers ln the rtght-hend column
tndtcete Ibe next stege et whlcb. Iirst. the named unit
and. second. ita cnmplement. are encountered. Thus it
mllk.es It possiblete tracethrougb this aecountthe course
of disarticulatlon of a particular par! of Ihe skelelon,
such aathe forelimb... lHiII1979:741].
The hasic sequence of disBrticuiation is head and
frot leg, followed by rear 188, and finally the ele--
menta of Ihe axial skeleton-. 'On the basis of these
observations, we mighl expect the following anatom-
ical segments lo have slightly ndependent distri
butions under nalural conditions simply as a func-
tion of the sequence of likely disarticulation.
1. Cranium plus alias vertebra
2. Fronl leg
3. Rear leg
4. Basic axial skeleton (cervical, thoracic, lumbar
vertebrae: sacrum, pelvis, ribs]
Far inslance, if one can picture competilion
among camivores arDllnd a dead animal, then it
might well be reasonable to anticipate that skull.
mandible. 8nd front leR parls would be most Tikely
dragged away from Ihe kili {rst during the early
period of competition far parts of Ihe kili amang
numbers of
In addilion, we may anlicipate that parts w[1
"ride" wilh others as a funclion of their probable
lttle "meat," as is the case with ungulete lower
limbs. In a contrary fasbion, where there s a concen-
Iration of decomposable tssue. as in the abdominal
aree. chemcel reecttons essocteted with decomposl-
on will hasten dsartculeton of such perts as rbs.
The second study of dsertculeton is one con-
ducted by Andrew Hill (1975). This investigalor ac-
lually dd a pattern-recognition study based on the
comperetve degrees of disarticulation manifest at a
series of deeth sites of DamaJiscus kcrrtgum, or topi.
The average adult body weighl of topi is 82 kg or
about 180 lb. His findings on the eequence of dser-
liculation are as follows:
Fromthese observations (he normal coune of dllUlnicu-
lation can besummarized approximately aa fol1ows: (1)
skull and sorne limbs become disconnacled {ponibly
also the atlas), {2) tlbll become loosened and fal! off
at leasl from the upper side. [3) Ilmbs stan lo disar-
ticulate Into progressively smaller segrJHlnts unll only
isolaled bones are lefl [sometimeduring this process lhe
lower jawbecomes disconnecled 1,(4) follawing this
bul overlapping jls late st8ges, the vertebral calumo
S18rts lo dlsarticulate, and (5)afler this haacontinued far
sorne lime bul befare diSArliculatian of lhe vertebral
column s completed, weathering. ,plinterng and
gradual dlslnlegratian of the bones set in [Toals
It is interesting thal Tools attributes Ihe resislence of
Ihe vertebrae to Ihe inlerlocked Iype of articulation
and the rapid disarticulation of joints such as Ihe
atlas-axis lo their highly mobile character. He fur-
ther notes that hide and ligaments retard disarlicula-
tion and this Is particularly lrue where desication
occurs befare decay by virtue of there being very
Skeletal Disartieulation
There heve been very few studes of natural
skeletal disarticulation. The two that are available
are somewhat ambiguous on whether the active
agente are decay organsme end chemical processes
or whether there has been ecton by predator-
scavengers. In the earlier study eveeble to me
(Toots 1965), it would appeer that the agente of ds-
articulation are primarily decay organisms and
chemical processes. Toots summarizes his observa
tions as follows:
spiral fracture when produced by anmele Is not lm-
ted to BO origio at the distal ends of the bones. [d]
tbe inodifications citad by Frison al Glenrock as ev-
dance of muscle stripping are cornmonly produced
by gnawlng cends. (e) the modcatons citad as
criteria for recognizing "expediency tools" are
cornmonly produced by nonhominid predetor-
scevengers. and (f) many of the modcettons pra
vously citad as evidence for human modcaton are
'-referable to predator-scavengers.
c",/.",.Jq-, /"k-o
3. Paltems of Bone Modi/icarions Produced by Nonhuman Agents 44
wolves had dlsmembered thetr prey by (a] eating
thl'EH.lgh tha proximal humerus, therefore dteer-
ficulatiog the scepula. which W8S Ihen deegged off
(the otherwtse unmoded front leg frcm the mddle
of the humeros shaft down remotned ettached to the
body by sktn in two cases sud WBS scattered in ntne
cases), and lb) cracking the fernur in roughly md-
"haft so that the proximal end of the femur remened
);lfliculated lo the pelvis (however, tha greater Ira-
chantar was eeten away]; the lower rear lag was mis-
sing or wdely dispersed around the k.iIl lncation.
Mueh lesa common was a break in the Jower thrd of
the libial sheet such tbet the complete lower- leg from
the distal third of the tibia clown remaned artieu-
Jalad, sornetimes with pert or a11 of the toes and
hooves mesng. This meana that the proximal femur
and the proximal tibia may remaln essoceterl with
the pelvis. the distal fmur end the proximal tibia
may remen arttculeted. end the distal tibia clown to
the foot represente the final segment of the rear leg
typico.lly found artlculatad. In the case of the front
leg. the most comrnon pettem wes for the scepule to
cccut sngly, the proximal end of the humeros to be
destrcyed. and the remaining parts of the front leg to
remain articuiated. An altl!'rnativp waa fm there to be
a spiral break through the lower part of the radio-
cubitus so that the distal radius was articulated with
the carpals, metae8rpaJ. and segments of Ihe fool.
Another common segment comisted of the main
body oC the shaft o the radio-cubitus, and tha distal
humerus, plus a considerable section of shaft.
1greatly regret that I did not OJllecl all the bones
from the WCllf kills for re'llm to Albuquerque. Al fhe
time 1 was intetested in anatomical part survival and
did not make systematic notes on either units found
in artkulation or the dstribution o tooth marks.,
fracture patlF!rns, or nther evidence of carnivote ac-
tivi/y. 1 saved only bolles exhibilinB extraordinary
evidence of earnivote destruetion. Gary Hsynes
(1977,1978&, 1980bj has informed me that he has cur-
rently underway sn extensive study of wolf kills oC
American bison. He expressed the apinion that Ihe
paltern o( dismemberment was so regular tha! we
might eventually be able to tell how hungry the
werF! snd the size of the Rroup making the
kilI simply CromIhe patterns of disarticulation and
bone dispersion. Thus far, the consistency between
my observations and those made 011 ve!")' different
mimal!> in a very differenl would tcnd lo
support Ihe idea oC extreme ragulanty end. in tum,
ndcate that we mght well place sorne confidente
in making unlfcrmttenan assumpttons ebout pat.
terne of dismemberment when eontemporary pat-
terns are documented more fully.
Toolh Mar.ks
Viewing animals as denticulatec( is obvr-
ous Ihal the lmpiements that ultlmatety modify
borres are the teeth. Ignoring for the rncment break-
age end actual modcattons in bone morphology. J
will concentrete on surtrcal scarring caused by the
moton of animal teeth on and over the surfece of
J recognze four baste types of tooth marking on
bone: (o) punctures, (b) pits, le) scores. and (d) fur-
rows. These may be coupled with types of breakaga
and when distinctive, this will be mentoned.
Punctures (Figure 3.01} are simply where the
bone has collapsed under the tooth. frequently leev-
ng a fairly olear imprint of the tooth. as in the dorsal
[rght-ade] view of the distal end of the metacarpsJ
shown in Figure 3.02. Fraquently, when the bone ls
thin and/or porous, the tooth may penetrare the bone
leaving distinctive hales in eaneellous bone [Figures
3.01 and 3.36). On very thin bone, such as Ihe blade
oC the scapula, the tooth may penetrate and remove
an area of the edge equal to tne surface area of the
tooth, producing a crenulated edge (well iIIustcaled
in figure 3.40). This effeet has beeo illustrated pre-
viously by Bonnichsen (1973:19. Plate 3, item A.j
and Shipman and Phillips (1976:171, Figure 1, item
Punelures were one ol the lrst mod(caftons
made lo bones noted by flarly archaeologists. In fae!,
BOllcher de Pecthes. the father of modP-rnprehistory,
reported a punctured phalange flOm the site oC Ah.
beville (1849:312). Punctured bones were noled wlth
inleres! b}' Larte! and erisly and a considerable dis
cussion in Ihe Freneh literalure lollowed the sugges
Iion Iha! these were whistles, Ihe first musical in-
slrumen!s. The Iiteralure is well summarized by
Henri Martn [1907-t9HJ:162-168), one of Ihe firsl
to point out Ihar perforaliulls were commonly made
on Ihe weakesl surface of abone and to givfJ many
FIGURE 3.01. PuncluM& mude by (lnJmol 16flth,
FIGURE 3.02. Animal-prodllced pitting and punclures on lhe distal metapodial.
! li
'ni I
" l'
1:. 1:1'
'!! !
1, jlll
1II !
I 1I
:f: 1i'
1I '
eA-? <O.z..S_
-c <\ cOiY "
f?/ - l-i
3. Pcttems of Bone Modifkolions Produced bv Ncnhumun J\genls Tcoth Morks
FIGURE 3.04. Animai-produced pitted and scored com-
pacl bone sulface.
'" I i!,"
1,' '
ji /:1
1"1' :
. }
: I

1::1: I:,,'
and furrowing. The bone is generally so enough so
thet scorng rarely resulta from normal gnawiIlg, al-
though il may occur on herder edjacent parts of the
bone. FilIeting generelly results in cut marks parellel
to the long axis of the bone, whether long bones or
flat peces sueh as the scapula are nvolved. Scoring
Is almos! witnout excepuon transversa lo the long
axis of bcnes. ence it results from turnng the bone
egetnet tha teeth ralher then "pullng'' the teeth
elong the bone (Figure 3.06). Actuelly, aer sorne
expertenca with cut merks end scorng. I found few
embguous si/uations, except perhaps where pups
had been feeding and their fine pinpoint teath pro-
duced a very sharp seoring, sometimes even Ion.
gitudinal scortng from tugging and pulling in play
and competition for bones.
S[;oring has been well Jlulifnlted in the litera/ure.
In Breuil (l939) many licored bones from Choukou.
tien are ilIustraled (Plale XII, p. 65, Nos. 1.2,3; Plate
XXVI, p. 93, No. 5) afthough these are Interpreted by
Breuil as evidenee oC human workmanship. Pe
(1938) alsc iIIustrates items from Chou)outien but
recognizes them as the produCl ef animals, princi-
pally the hyena {Pei 1938: Plates X. XIII. XIV, and IV,
and XX, No. 2J. The foregoing are all good iHus-
trations of tooth scoring by relatively large clU'ni-
vares. Kitching {l963l iIlustrales numerous scoted
FIGURE 3.05. "CDlnpreSSur" pined and scared by ani-
MUR/e II 2["
111111111 1111111 i I
e ....fW
shor! curvature and shallower on surfaces with long
curvature or sections of bone tending toward fiat sur-
faces. Another majar difference ls the placement ef
cut marks versus seores on bones. Cut marks gener-
IIl1y resoli from three 8clivjlies: (a) skinning, (h) dis.
artieulation. and (e) filIeting. Cut marks from skin.
ning may occur around the shaft of lower legs and
pltalanges and parlicularly along the lower margins
af the mandlble ar on the skuIl. These are the only
marks that rnighl on oceasion mimic laoth .'Icariog.
Cul marks from disarllculation nOnDalIyoccur (a) on
Ihe edges, or articulator surfaees. al the ends oflong
bones and (b) on the surfaces of vertebrae or pelvie
parlt;.These are all areas af relatively soft caneellous
bone and ere ept to exhibit puncture marks, piUing, FIGURE 3.03. AnimcJl.produced cJdrmsive/y pitted bVDe.
rendering lt unlikely thal compressors were used in
the suggested manner. Nevertheless, Bordes eccepts
these items as tools wth the warning thet they were
more than Iikely scerred while being used in percus-
sto , heve examined many of Ihese items
(approximateiy 134 from tbe slte of Combe Grenal)
and in most cases Ihere is no tendency or thetscer-
ringla be on the end where one would expect il if
these toms were used as flakerl;. In far:t.l could see
no relettonshp between the placement of pilting
(sometimes coupled with scoring as in Figure 3.D4)
and the rnorphologv of the spltnter. lt was as if the
spttnter was produced after the pitting had been
These ilems have been widely ilIustrated as Iools.
by Bordes (1961:Figure 108, 1968:100, 111, Figures
34 and 39), Semenol' (1964:171-172, 177. Figure 92),
Movius (1953:45, figure 12}, and DeLumley
(1969:165. Figure 20), among mllny others.
Senrin! is the result 01 either lurning the bone
against the teeth or dragging the teelh aeross rele-
tively campact bone. The result ls a scarring of tite
surface that IScommonly linear. Scodng may on oc-
casion look vory mllch Iike "cut marh" from lItnne
tools in that scoring marks may be parellel snd fre-
quently quite clase together. They generally differ
from cut marks in Ihat cut marks rarely follow the
"eonteur" of Ihe bone, beinS deeper on lhe par!s of
examples tdentical lo that illustrated here {Figure
3.01. upper speeimen) (see Martin 1907-1910: Platea
SiRCepunctures ara a fundan of the strength o
the bone relettve lo thc strangth of the animal, ! s
not surprsng lo nd large end deep puncture marks
made by lions and other larga cats. [See, for nsence,
Bonnchsen 1973:21, Plate 5, ttems A and B; Dert
1949:13; Hill 1980: Figure 6.4; Kitchlng, 1963: Plato
.. ,: Pei 1938: 62. Plata XX!, Iteme 7,8,15, and 16.)
Puncture marks are really the initial steges of
furrowin,R-8. tarm edopted From Haynes (1980). That
te, Ifconsiderable gnawng occurs.ubvtoue puncture
marks are generally oblterated as the cancellous tte-
sue ls gnewed ewev. Gnawing ganerally proceeds
foom 10ft to harel bono; the animal attacks the smt
cancellaus plU1:s af abone first. As it gnaws, it en-
counters progressively harder bone {Ihe bone shaftsJ
and the same aeUans may result in pitting in that tbe
bone ill now strong llnough not to coIlapRe under the
gnawing acHon, This pitting ls ilIustrated in Figure
3.02; the bone on the left is ex:tensively pitted,
whereas the one on the rlght is ptmclured. Figure
3.03 illustrates in more detaJl the effects oC extensiva
pittlog. This is a cemmen conditon on bones that
have been gnawed extensively. I should poiot out
that I have observed this only on bones from wolf
dens and dog yards. No exomples e bones so exlon-
sively gnawed were observed on the wolf kills that 1
will discuss Jaler. PiUiog generally results from
gnawing bones rather lhan eatins and pulling meat
from the skeleton of an animal, as at a kili. Bone
gnawing as suth is rlU'e al Idlls but very common at
dens and the "sleeping" places o canids that 1have
observed.1l ise even less likely at the killsites of large
cats(Diane Gfford, personal cornmunica(ionl.
Figures 3.04 and 3.05 illustrate Ihe pitted surfllee
of a rompacl bone splinler. which if found in Ihe
European Middle Paleolithic would be called a
1001 and designated a "campressor." This category
was originally described by Martn 0907-1910),
who Interpreted it as a campressor or pressure nakee
usad in stone 1001 manufacture. The pitting and scoc-
ing of Ihe bone was viewed as Ihe resuIt of pushing
off flakes of chert during the ael of pressure flahng.
Bordes 119Bl :77) has correclly pointed out thllt the
very presence of the technique af pressure flaking
has nol been established for the Middle Paleolilhic.
f"''''V- J c-&J&
'.' "

48 J. Potlerns of Bone Modificolioru Produced by Nonhumon Agenls

OtherSeu-ces of SUljiciol Modificotions 00 90ne 49
FIGURE 3.06. Compile' bone scored by animal leelh.
bones from what was mas! cerlainly a majar hyena
occupation of Pin Hole Cave in England (Plates 5, 6,
7,15). MilIer [1969: Figure 3, hottom, 8nd Figure 4,
top) providas gond illuslralions oC tooth-scored
bones. as does Haynes (1980: r'igure 1 sud 2j.
.Furrowing Is a term 1have adopted from Haynes
(1960) lo refer lo Ihe effect Iha! repealed jaw Belioo
with either canines or carnassials produces on reJa-
tively cancellous bone tissue. The result in extreme
cases is what Bonnichsen (1973:16), following
Sulc/ffe (1970:112). has called "scooping out." In
furrowing soft Iissue is removed. leaving 8 "hole"
which is graduated up agaios! harder, mOfe compacl
bone. The aetion ofthe leeth will sometimes produce
undulations on the remaining surface of the eanee!l-
ous tissue, and there are sometimes steplike ir-
regularities where the teeth have vised down in dif-
ferent spols as the bone is mouthed by the animal.
This is by far the most common type of damage pro
duced by animals and is the lype thal is presenl even
ir animals have given ooly minar allenlion lo Ihe
remains under investigation. Examples offuITowing
are cammon in Ihe Iiterature (Bonnichsen, 1973:
Plates 6A, 6B, 6D; Haynes 1980: Figures 3, 4; ;MilIer
1969: Figure 5; Pei 1938: Plates 12, 20; Sulcliffe
1970;). In the older lileulure, a number of authors
have interpreled furrowing as Ihe conseQuence of
human modificalion (Breuil 1939: Plate XV, No. 2,
8nd Plate XVI, Nos. 5 and 6; Weidenreich 1939: Plate
VIII). In Ihe more recenl literature furrowed pieces
have 1meninterpreted as Ihe consequences o( butch-
eriog wilh choppers and "musde slripping" (illus-
trated in Frison 1970: Plates 9, 12; Iohnson 1977:
figure 11.2, to rnentton only two).
Figures 3.42 and 3.43 nicely iIIuslrate furrowing
of cancellous tssue by dogs and by wolves.
I made eeverel observettons that may aid in die.
linguishing the behavioral context of different types .
of scarring by animals. Wolf kills predominantiy
ylelded sorne furrowing, relatively common pune-
ture marks, and sorne crenueteted edges: pitting
and scoring were much less common. On the other
hand, the Eskimo dcg yard assemblages end the wolf
den assembeges yielded extensiva pitling, scoeng.
and more extreme furrowrng. This seems to reeult
from the "boredom" chewing of bones in the dens
and the dog yerds: meat consumptlon was primarily
al the kills. At kills, furrowed destructon of soft can-
cellous tissue while meat te still present on the shafts
of bones prevente Ihe harder bone adjacenl to Ihe
cencellous ttssue from bsoomlng heavily scored or
pttted. as s ilIustrated in Figure 3.43. Stated another
way, when rneet is present in considerable quan-
tities, long bones may be furrowed and even
pollshed along the graduated edges wilhout corre-
lated puncture marke. plts. or heevy scoring on edja-
cenl denser bone. Thls may be particular/y true
"",hen animals are altempling lo dismember a car-
cass, as at a kili.
Dlher Sounes of
Surficial Modificatiohs un Bone
lo my discussions Ihus far l have only addressed
the issue ofCarnivora versus man as an agenl oEbone
modi(ication and surfidal scarriog. As mighl be
imagined, Ihere are olher agents thal may syslemati-
caUysear bones aod othar na/ural conlexls in which
bones may become modified.
Perhaps the most remarkable of Ihe agents whose
modifications of bones may he misfaken for human
workmanship is Ihe porcupine. Pe (1938) ilIuslrales
numernus bones morlified by Ihe porcupine and
olher rodents. Rone ""hisels" reported (rom
Hopefield {Singer H15fil. and modified bones from
ancienl rlepnsils in China (Hooijer H151). although
presenled as of human modifkalion, ap-
pear to have been modified by rodents. The
Hopefield bones are particularly interesting, since
Stnger reports en essemblege of "control" specimens
recovered from a roek crevce considered too small
and low to have been used by manoGiven such con-
dtttons. il was reasoned that the bones within could
only have been introduced by entmele. The control
assemblage was then competed lo the so-called
chisels from Hopefield, and the conclustcn was
ruede Ihat animals were Indeed responsible for the
modtcettons on the Hopefield specmens.
Lesa settsfactcry claims accompany the studies
conducted al Kalkbank (R. [. Mason et al. 1958), 8S
well as tbe Israeli Mousterten cave ot Geula (Oart
1967). In both cases rodente had modified the bonee
constderablv: nevertheless cloims that humana had
mcdted them were made. Even more remarkable is
the fact that the Kalkbank rnatertel is consdered by
Darl and others as a control essemblege providing
information on the diagnostc propertes of bone as-
sembleges modted by mano No less blzarre c1aims
are made by Dorl (1967) for the Geula meteriale.
very charectertsttc patterns of gnewng iuvolvng
parallel tooth merks. productlon of "windows" in
the shafts of bones, and extensiva modifications in
localized areas, particularly along breaks, are all
diagnostic properties. Tha relevanl ilIustrations pre-
sentad in Pei (1938), Singer {1956), Oarl (1967), and
Bonnichseo (1979) should be consulted for further
As in the case of Ihe rodenl modifications, Pei
(1938) was one of the first to recognize Ihat consider-
able alterations 00 bone surfaces could result from
Ihe solulion of bone surfaees in sorne situations
where dense roots covered Ihe bone. Such etching of
bone surfaces in conjunction with Ihe presence of
plant roots is common in sorne environments (see
Figures 3.07 and 3.08). Human agency for mod-
ifications most Jikely referable to lhe Jatter con-
ditions has been c1aimed by Bordes (1969) for an
"engraved" bone recovered from the Acheulian
levels of lhe site of Pech de L'Aze, and by Freeman
(1971) for a piece of "Arte Mobiliar" from
Mouslerian Level 17 at Ihe Spanish cave of Morin.
80lh bones are des"ribed as havinJj\ a decorative
"macaroni" Iype molif (Marshack 1976). Bordes'
find has been c1aimed to be Ihe earliest evidence of
man's symhoJizing and arlislic ca.pacities: "If my an-
1" ;11
;; ,
r' i ,

,,j ,
".,, 1 ',1
.'I '
"f '11'
'j'" i j
i J.l
l' 1

i! 1

Y "
FIGURE 3.07. "Arle Mobilior" produced by odds in compact rool mo.ues on Il' jowfrom 'he Beur sile, Anokluvuk
FIGURE 3.08. nelail o/ bOIUl in fiRure .1.07.
acne Brt!!okolle and Desteuetion by AnimaJs
elytc resulte Ion Bordes' spectmenl are correct, tbey
may change our concept not only of the ortgns of ert
and symbol making. but also of the intellectual and
cultural evolutton of Horno scprens [Marshack 1975:
85J," Clatms such as these are certair..ly the basts for
modern myths ebout ancient mano
Bone Breakage and Destruction by Animals
Many hours o observatlon of both wolves and
dogs revealed a very convtnctng and redundan! bone
consumplion strategy, which lea ves a regular pattern
in the Faunal hy-products of animal consumpton.
Gnawing animaJs aliad. the ends of long.bones first.
They are moet successful in chewing away the sef-
teet.ertculaec.end. The softer the articulator end the
faster the animals chew them away and expose the
cross section of the medullary cevtv. At this point
en animal may lick at the end end "scoop out" mar-
row by turning"'TIie canines around in the cavily and
simultaneously Iicking into the cavity. but this is
very unsalisfaetory. More commonly the animals at-
tempt to collapse the bone shaft by use of the strong
cheek teeth.
Given a very resilient bone. the animal rnost
commonlv turns it parallel lo the orentation of the
camesstal teeth and altempls lo puncture the crown
of the bone eylinder. which results in what I call
cnonneJed bones (see Figure 3.09). This effeet is
produeed by puncturing the bone baek from Ihe
transverse edge. leaving a channel running parallel
to tha longitudinal axis of the bone. The animal may
proceed in Ihis manner for considerable distance
"clown" a bone. In Figure 3.09 Ihe upper bone is the
shaft of a long hone, which is c1early scored and
chipped along ils edges; the channel is evident. In
Ihe lower example. a dorsal spine of a thoracic ver-
tebra, Ihe punclure marks are c1ear. Channeling is
unique to animal gnawing and in my experience
there are no human proeessing lechniques Iha! result
in analogous forms Isee BudJann 1823:276; Sutcliffe
1973:Figures 2. 3. 5; Zapfe 1939:1511
In addition lo lhe "rlirer:l" bi!tl-duwn approaeh.
which is mosl common when a msislant bone is en
eounteren.lhe animal may dslurl its face, moving ils
jaw lalerally away from the stde where the bone is
tnserted in the mouth, and tum its head into the
bone remcvtng a splinter along the shaft. Success in
ths strategy results in the removal of a substantial
ake parallel to the longitudinal axis of the bone,
expostng more of the marrow eavity. Such a fleke is
generelly ponted and scarred; the opposite end
commonly shows "step fractures" (Figure 3.10). This
paltern ts known from many locetons of ercheeolog-
cel inlerest and s commonly interpretad as bone
modficaticn by man {Bonnichsan 1978:112, Figures
6a, 6b; Frison 1974:42-43; M. D. Leakey 1971:Figure
36: Stanford 1979a:108. Figures 8, 7). Such fleke re-
moval originating at the broken end of a bone may be
coupled wth polsh as in Figure 3.10. or radel scar-
ring as in Figure 3.11, and also has been teken as
evdence of tool USe. (See Kitching [1963: Figures
5,6 7bl and also Klima et 01. (1962:150. 154. 155.
1571 for examples of radial scarring from a hyena
Another characterlstic form of behavior ls what I
call "chipping beck" an edge. This s done by using
the strong carnassial teeth. Abone is placed in the
mouth parallel to the orientation of the row of teeth
and a ridge or protruson s placed between the car-
nassals and "mashed off." When the bone s dense
compact bone. the result of this viselke pressure is
an effed very much akn to preesure flaking (Figure
3.12). Small chips are pushed off, resulting in an
edge along a break that may be continuously
cbpped. showng all the features of a "microden
tieuIate" in Iithic terminology (see Figures 3.13 and
3.14). Frequeotly associated wlth this characteristi-
cally microchipped edge is tooth scoring of the ex-
ternal surfaee of the bone below the chipped edge,
As the animal chews on the exposed edge. chips are
driven off. When Ihe edge gives way, the large teeth
slide along the surface of the bone, resulting in typl-
cally parallel snd slightly oblique scars below tbe
chipped edge. This is well ilIustrated in Figures 3.15
snd 3.16. Channeling is often combined with chip-
ping baek to produce very convincing pseudotools
(Figure 3.17),
Chipped edRes as iIIustrated in Figure 3.12 are
almost inevitably interpreted by arehaeologists as re
touch. Dart (1960:5} ilJuslrated eight ehipped-back
nekes from grey breecia at Makapansgat, about
which he commenls: "He also prepared feeding
" \
1': i
l' "
1, ,
, .
I i i
f I
i: i I
y 1
i: Il!!
".; ,
:,;"!I: f
.: .',
" 1
- .",
;;"..... "._,,-
FIGURE 3.14. Chipped.bacJe eriges showlng mlcroden-
'lculored effm:t.
FIGURE 3.15. Edemal face 01chipped.back bone sho....-
lng obllque roolh scamo,.
FIGURE 3.16. External fucf! o{ chippoo-back bOlle 5ho.... -
inS 'ooth scurr;ng mm "slipplng down" lhe buRe K-';Ih
lhe teetb.
FIGURE 3.12. "PressuTe f1oking" un lhe end of a gaawed
bone resulling ram chipplog back Ihe edge.
fiGURE 3.13. f....rlensively f'hippcd oock ['dges uf
lIllawed bone.
flG\JRE :1,11. Pitled und radial/y scarred b(me, the re-
5u/1 of extensive goawing.
3. Putterns of Done Modi!lcalions Produced by Nonhuman Agenls
FIGURE 3.09. "Chunne/ed" breokoSe of bones by anima/s.
FIGURE 3.10. 1IighIy ptlllshed end of a gnawed bone:

- (.,UA-{;;....) l
e:' -'-< ti/, ---------=Atr.-'---------

J. Pcnems of Bone Modificotions Produced by Nonhuman Agenls Bone Breokoge ond Destrucon by Animols
FIGURE 3.17. EKomples o{ chunne/ed ami chipped-back bones cornmon/y c01Ifused wllh humon workmanship.
utensUs aud he split bones longitudinally by percus-
sion. He sharpened and pointed by trimming either
the broken shaft tseU or Ihe flakes derived froro
splitting Ihe bones. He devised and prepared his
lools IOart 1960;1371." Needless lo say, the "he" is
Austrolopithecus. For soroeaDe accustomed lo work
in lithics. 3n encounter wilh a chipped-back bone
produced by animal:! will almosl always convince
Ihe observer thal he is seeing Ihe work of mano For
instance, very salid ohservers of artifacls such as
Veyrier and Cambier (1952: 383-385) announced
thal they had confirmed lhe manufacture of bone
tools during Ihe Lower Paleolithic, illustralinR what
are almost certainly bone flakes chipped back by
animals. Earlier researchers had been equal1y im-
pressed. For instance, Breuil (1939) presents numer-
ous examples of chipped-back bone flakes as evj
dance ror the bone 1001 industry al1egedly produced
by Sinonlhropus at the site of Choukoutien (Breuil
1939:Plates XVlI1, No. 3, Plate XXII, Nos. 5,6, lIJ. 19,
plate XXV, Nos. 9, 11, 131. Breuil stales,
Long bones of a1l kinds. chiefly humerus lIfldcannon-
bones. were Irimmed al !tll1 nreak as chisels or poinls,
Ihe convex distal articulatioo !nrmiog 11 handle: from
Cnoukoulien to historie times Flakesmay have beeo
used just as lhey wera. or trimmed at the ends (cheels.
percers. or points) cr trirnmed along the sides [scrap-
era]. Plekes of verv big bones were used after being
trimmed like nint' n.u.es of the same size and shepe
(Choukoulien. Jower levels of Castillo includlng the
Mousterian) [Breui11939:6-71.
Recently DeLumley (1969) reportee retouched
bones representing 13% of the total tool Industry
from the Upper Acheulian ste of Lazaret. Chipped-
beck spectmens are well iIlustrated in Figure 17
\1969:121. where tems 3, 4, 5, 6, and 7 are extremelv
good examples. Figure 18 (1969:164), item 1, s a fine
example thet also exhibits the frequently correlated
presence o both pitting and sconng. It should be
poinled out that wolf rematns were found in the
cave, as wera coprolites of predators. 1 am confident
that there was at least a minar occupation of Lazaret
by wolves. Charactenstically. ther remains have
been interpreted as the result of human behavior.
In New World studies two chlpped-back frag-
mente. both desgneted ertlfacts, are illustrated by
Bonnichsen (1979) from the Old Crow Flats collec-
tions. One is called a "bone core" and the second is
designated a "mtscelleneous polished and ground
bone artifact" (Bonnlchsen 1979: Plate VIlI-8, and
plate VIII-24).
The most enthusiastic recent advocate of "re-
touched" bone flake tools in the Paleolithic s Leslie
Freeman, who eomments on his analysis of such
"tools" from Level 17 at the cave oCMorin: "If the
bone pieces were omilted from the study of recavo
ered artifacts, an impressive richness of detail con-
ceming the teehnology of the Mouslerian cave oc-
cupants would obviously be los! (FreeIDan
Wilhout taking them up pieee by piece, 1wiH as-
sert that the chipped pieces iIIustraled by Fraeman
(1971, 1978) are c1assic examples of chipped-baek
flakes produced by gnawing animals, probably
canids. Al! combinalions of telllale praperties are
represenled in the Morin Level 17 assemblage: (o)
chipping baek with diagonal seoring on the external
face of the flake. (h) ehipping back with pitling
and/or scorinR on rile same pieee, and (e) lypiealIy
denticulahnl eriges. Such combinalions are found
wilh pieces lhal are described as "eroded," as if by
the aelion of n:irl (Freeman 197B:4fi): Ihese are mosl
cerlainly bone frilgments passed in scal. It IS proba-
bly not wilhuul siRnificance Iha! both wolf snd
hyena were idantified {Altuna 1971:380-381}. Fl-
nelly. the "unique" charecter of Level 17 s com-
menled on by Altuna as folIows:
se nota una clara diferencia entre el nivel 17 (Mus-
lerense) y los nivels del paleohtcu Superior, no solo el
la compislclen faunis faunstjce sien tambin en la
composcton de las partes del esqueleto que eperancen.
Como hemos ndcedo un el comienzo de la parte gen-
eral, en el nivel 17 el munero de lI'5quirlas de dlafisis
excede con cucho al de epifisis que seria de esperar si no
hubiera habido una seleccon intencionada 11971:3851.
As will be shown In the fol1owing section, the
number of bone epltnters is inversely correlated with
the numberof erttculetor.ends in enmel-gnewed.ee-
semblages. Jusi the reverse is the case in well-
preserved assemblages of humanly modified bnne.
.'Y.Pt;l _9r
that animals are capable o On
moderately hard bone the animal may vise clown
wth ts teeth, at the same time ttltng lts head while
the opposite end of the bone rests on the ground
surface. This produces a double set cf stresses and
may result in the breaking off of an edge of the bone.
There may be no apparent tooth mpresstons in Ihe
break stde if the leverege is somewhal back of the
point where the teeth were Impactad. The edge pro-
duced in this manner has sorne of the properttes re-
sultng from mashng off the edge of a Styrofoam
eup, in Ihat there is a granular eharacter to the sur-
face of the break.
In lhe evenl thal a mashed edga is then licked
repeatedly, it may oblain a rounded polished ap-
pearance and may appear as "use wear" as illus-
traled in Figures 3.10 and 3.18. Such counded and
polished edges are regular elements in both dog yard
and wolf den assemblages. However, they are not
in a dog yard population of 866 bone
splinters and 267 articulator ends!
When a strong animal is chewing 00 a long bone,
there are always bipolar characteristics in that the
chewing occurs between the two sets of teeth. This
Crequently results in what has beeo previously de-
scribed es pitling. For instance, with the melalarsal
i1hlslraled in FiRure 3.17, when the dorsal edge of
Ihe bone has beeo chipped back, there are accom-
panying pits on the opposite side of the bone repre-
senling the impacl of the leeth serving as the anvil
fur Ihe action uf the opposile teeth. Such pilting is
1; !
:1" l'
.t ] ,
il.; I


r ...
J. Patlems of BoneModifications Pmduced by NonhumonAgenfs Bone Breokoge and Destrucfinn byAnimols
bone as has been suggested by Bonnchsen
I have watched dogs and wolves break caribou
bones in their mouths. The resulting splinters have
al1 the formal properties of the much discussed spl-
ral fracture. The bone splinters and the broken ends
of long bonea recovered from the dog yard samples
exhtbtt ell the formal chereoterstcs of spirally freo,
tured borre. As was poinled oul in my earlier dscus-
sion of these aamples, the long bones were broken by
dogs-Eskimos do not Ceed their dogs marrow-
crecked bona splinters. In the case of my controlled
dog-feeding record (see L. R. Btnford 1978b: 262-
265), there is no question as lo the stete of the bone at
the lime of its being fed 10 !he dogs. Long bones were
unbroken. On recovery. 44% of the long bones were
broken with a spiral fracture: Ihis was done by the
dogs. Broken bones and bone splinters recovered
from two wol dens (described in Chepter 5) charac-

HlI's coJleclions are areal samples in Ihal they
represenl aHIhe bones remaining on Ihe surface in a
given lUea. Mosl, if not aH, Ihe bone modification is
considered lo be the result of nalural death and sub-
sequent patterns of attrition by scavengers and
wealhering agenls. In mosl ceses the predator is be-
Iieved lo have been a large cat, Iion in particular.
Sinr:e Hill dnes nol know with absoluta oorlainl:.
appeers Iha!the Internal structure of the bona controla
the cbarecter of the break{Hill 1976:335; see etso HiII

FIGUKE :J.20. Long "loRSiludillully" split fragmenls with denticulaled ends.
Spiral frllclures are found lo occur liS1I result of nalural
weathp.ring and cllrnivure lIdivily lhal in Iife are sub-
;ec!p.d lo lorsional sll"f>ss, such liShumeri and libiae. 11
teristically exhbited spiral fractures. Broken bones
end bone spltnters recovered from the 24 separate
kills oE ceebou by wolves (lo be reported in Chapter
characlerislically exhibited spiroJ fractures.
My obseryations afe got unioue. Hill has re-
peatedly reported Ihal spiral fractures were consis-
tently noled in Ihe faunal coJiections me de by him in
Uganrla and Kenya:
f.ylinders with distim:/ife fi'(ilJRE 3.19. c"Jltlpsfld
FiGURE 3.18. Rounded edse on gnowf1d bone.
strtktng on the more cylndrical long bcnes of the
Iemur and the tibia.
Pitting occurs typically when an animal s 81-
tempting lo collapse a cyhnder: thet ts, it has chewed
off the articulator end and ts working 00 the cylnder
and the exposed rnarruw cevttv. Al Ihal juncture it s
no! uncommon for Ihe animal repeetedly to place the
cylinder al right angles to the carnassial teeth and to
vise clown hard in an ettempt lo collapse il. Jf the
bone ts dense, there s a puncturing or splitting on
the short are aide o the bone. Once this occurs Ihe
animal begtns chipping beck along the puncture or
split, and {he underside of Ihe bone oppostte this
split becomes tncreesngly pitted and heavily scar-
red. Finally, the animal may be successful in colleps-
ing the cylinder and ene ie left wilh the mpected
scarring on abone that has been split away from the
upper chipped surfece.
Under the latter conditions one gels very long,
Iongttudtnally spttt splnters wilh chfpped and
denliculaled edges along one or both ends. This
condition is iIluslraled in Figures 3.19 and 3.20,
where "worked" points and edges on exceplionally
long splinlers are shown. Ironically, what was
c1aimed as di.qtinclive of human behavior by Rreuil
(1938::18)tufOSout to be diagnostic of animalgnaw-
ing when we have conlrol material.
Menlion of splinters necessarily introduces
uf spirAI fractures and Ihe deKI.e-.to
Ihay <lre uniquel}' characterislic uf human
3. Pctterns of acne Modificolions Produced by Nonhurncn Agenls
1: 1I
' I
FIGURE 3.23. De'ojJ of poin' on o pseudotool shown in
Flsure 3.22.
FlGllRE 3.22. Pselldolools produced by rmlmols chewin8 o, ridses.
FIGURE 3.21. Group uf "ridlle-cresl removed" flokes
typieal of onimlll snawjnll
Feeding area
Not gnawed
(46 scals)
Size (cm) No. % No. % No.
0.0-0.8 37 .29 10 .22 O
10-1.9 29 .2J 9 .20 O
2.0-2.9 2J 16 8 .18 O
3.0-3.9 17 .13 8 .18 I
4.0---4.9 14 .11 9 .20
5.0-5.9 3 .02 1 .02 9
6.0-6.9 3 .02 O 13
7.0-7.9 1 .01 O 6
O O 1
9.0-9.9 O O 2
10.0-10.9 O O I
11.0-11.9 O O O
12.0-12.9 O O 2
13.0-13.9 O O 1
14.0-14.9 O O O
15.0-15.9 O O O
127 45 42
"See Chepler 5 IDr I descripliDn af Ibis 8ite
TABLE 3.02
She Dlstrlbuljon of Frogmenls Recovered fMm
lfiJcmalaiyak Wolf Den
contribute sprally fracturad bone lo the paleontolog-
leal record.
The ccnditon of splinters remaining at a wolf
feeding arca is of sorne interest. First, the smalt
spltnters o bone that are pressured off a bone by the
gnawing animal are most commcnly swallowed, re-
sulting in few chips end small bone fragmente re-
rnainlng in a feeding area. The latter. of course. ap-
pear in thescaC\...hich occurs concnntrated in ureas
of repeeteddefecatton around dens and rendeevous
loceuons. around marking spots. scattered peripher-
ally to ktlle, and scettered elcng tratls. Deteceton is a
common marking tecbruque of wolves and is fre-
quently done on prominent rocks eround an animel's
territory. I heve observed considerable bone rnncen-
Irations eround marking spots that result from re-
peated lcng-term rnarkng in the same place. Table
3.02 presente summary data on fragment staes racov,
erad from the deeceuon and feedng areas around a
woIf den. An attempt was made to recover every tiny
fragment of bone in both instances.
agente responsihle for the bones. those who wish lo
in the equaton o man end spiral fractures
tend to dismiss Hill's data as "just a modern sample"
where tbe agent must be inferred as in any aro
chaeologtcal stuetton. This sama crtcsm may be
leveled al spiral fractures recordad by Shipman and
Phillips (1976. 1977), who coected areal faunal
semples from the Awash area ofEthiopia. In both these
cases the bear and the footprint are not, strictly
speektng. together. hui it s the optntcn of the re-
seerchers that the OOllJ' is not far away. They know
the reas in which they worked and find Hule reason
lo suspect that man was Involved in any significanl
manner as 8 causal contributor to the bone as-
sembleges collected.
Zoo studies, wbere the animals ere contained end
the parts fed to the anmals are known, are excellent
experimental oonditions. There ts no chance that
any agent other tban the animal fed in the given cege
oould modify the bone alter tt was ntroduced. This
strategy WIlS used by Dean Buck1and (1623). Pei
(1938), Zapfe (19391, end Bonnicbsen (1973), to meno
tion only a few. A current study of this type has been
under way for severa) years by Gery Heynes. a
greduete student at Catholic University of America
(Haynes 1978a, 1978b, 19801.
Paleontological studies-that is, the study of frac-
ture patterns from assemblages where roan could not
have been a contributing agent-are of utmost im
portance. ODe such study (Myers el 01. 1980) is most
In this study five New World assemblages of
fauna were studiad. Four are dated to more than
500,000 yeers ago; (he olclest, to 17 million years.
Man could not have played a role in the formahon or
modification of these assemblages. Spiral fractures
were common in these paleontological faunas. The
researchers further noted thal evidence of camivore
aclivity was not wel1 documented in Ihe form of
looth marks aud other telltale modifications and
ooncluded Iha! still another activity of animals,
trampling, is apt to have been the cause. I will return
to this point later. Of importance allhis poinl is only
thal substantial spiral fracturing is iIIustraled from
tbese localities and no human could have possibly
been responsible.
The evidence appears conclusive Ihal animals.
both through usiog their jaws and perhaps by tram-
pling the skeletal remains of olher animals. regularly
r UC'
Illl 3. Pcttems vf BOlle Modifil"()fivlls Producerl by Nonhumnn Agents
Modificolion5 by Allotomic()/ Purt
Several pettems of nterest eppear. Ftrst. the vest
mejorlty of splintera sud chips observed in scets
were under 4.9 cm in length. Basically the eeme size
range W8S observed in ungnawed chips and splin-
ters. 00 the other hand, fragmente showmg evidence
cf gnawing are certainly aH 011 Ihe large size o the
distrtbution. This makes considerable sense, stnce
the small flakes and chips are pressured off by the
gnawing animal duriog its reduction o Ihe bone,
end are generally swallowed. Elements oC the bone
that show considerable evidence of gnewlng, wben
they are broken. preserve areas of contguous chip
removal end scamng. These are larga splinters and
therefore will not be swallowed. Thus. when ODe ls
dealing wth an anlmal-gnawed assemblage. evt-
dence al gnewlng, preesure-Ileked edges. incised
scerrng on the outside of the Flake. ptting and abra-
sion Irom repeeted viselike mesbtng cf abone sur-
Cace. end so on will occur on lerge splinters (greater
than 4.0 cm in length]: small spllnters will exhtbtt no
such modtcatton. When small chips and splinters
are present in large numbers.tbey may sbow stgns of
hevtng been eroded by stomech ecds Iand (hus de-
posited in feces). (See Kitching [1963:19-22] for a
good descripUon of the effed of slomach adds on
booe splinters.)
One addilional characteristic. which appears to
be diagnostic of anirnal.gnawed assemblages. was
noled. Animals chew at protrusions of "thin" ridges:
8 bone is placed in Ihe mouth and irregularilies are
vised belween the camassials. This action results in
pinched-off tuberoiiities ando more importanl. in
long splinters lhat are basically triangular in cross
secUon. oflen with poinled ends. These splinters de
rive fromtbe biting off of ridges such as the pectoral
ridge. the supercondyloid ridge on the humeros. the
Ubial crest. the linea aspea on the femur. the ventral
ridges on lhe metapodals, and tbe ventral crest of
Ihe cubitus (ulna). When sucb ridges or crests break.
they are generally too long lo be swallowed and re-
main as very common poinled "pseudotools." (See
Figures 3.21, 3.22, Bnd 3.23.) Forly-three percen! of
the gnawed splinters recovered from Ihe den area
(Table 3.02) were of this type. 1 have never observed
such splinlers in percussion-fraclured assemblages;
percussion fractures generally cross such ridges
rather than run parallel lo Ihem. Fine examples af
very convincing "tools" are iIlustrated in Figures
3.21,3.22, and 3.23.
Modifications by Anatomical Part
As pointed out eerller. one of the more common
argurnents offered regarding the possible use of
bones as lools has been based on the "use potentlal"
of the shapes of varous bones. or their shapes as
modified through breakage, andlor citations of
worked edges ut SUMacas in conjunctlon with either
normal or breakage shapes. 1heve descrfbed in sorne
detall the types of edge and surface modrcetions
anrnals are capable of producing during normal
bone gnawng. as well as their distiuctlve breakage
1 will now describe the charectertsttc patterns of
destrucon end rnodificatton 1 have observad on
cerbou end sheep bones from the Esktmo dog verde,
as well as from the wolf denso In each case where
false or suspect 1001 Identflcations have been made
of similar perts. l will ettempt lo idenlify the relevant
literatura and cite the problema 1see with the clame
for human modcetton.
1have alreedy dernonstrated Ihat gnawng results
in distinctive patterns of scarring and chipping.
These condilions are aS50ciated with the pattern of
breakage. which has also bean shown to be dislinc-
tive and diagnostic of animal-modified assemblages.
Given Ihese general conditioos, iI is not surprising
Ihal there are distinctive patterns al deslrucUon as
sociated with the different bones as a funelion of
Iheir differences in slrength and overall fonn. The
following are observations of patterning in the
morphology resulting from animal feeding and
Examples ranged (rom a complete and unmodi
fied skull lo a skull of which only two parts
remained-the base around Ihe anllers (Figure 3.24)
and the palate with two looth arcs {Figure 3.25}.
Animals appear to begin gnawing from the nose in-
ward, removing the face and finally coJlspsing the
cranium. }eaving the palale and occipul in Iwo parts.
Rarely do the latter two parts exhibit further gnaw
ing. When the sKull is eaten in the manner described.
few skull fragmenta rernain at lhe localion of con-
sumption; lhey appear lo have been ingested.
FIGURE 3.24. Anlma/.,nawed cronlum witlr ottachetl
antlers (carlbouj.
FIGURE J . 2 ~ . Animul-NnUwl'd IIO/fl/r.s Icoribllu) shnwing t!istindivc chrJnlmling and crr.nu/o/ion 01 'hin bane.
l' '
;i 1:
'1'1 t i l ~ 1
[1 "
1\ r
!i ,
~ 11
T S" ,
3. Potlerns of Bolle Modificotiolls Pmdu<;;edby Nonhumon Agents Modificotions by Anatom(;aJ Purt 83
FI(iIIRt; 3.28. uf curibuu mandibu/flr deslruction by bo/h dogs (Iuwer two) ond wolves (upper lour}.
" I
dible s attacked by animals s partly a function of
differences in mandibular morphology among dif-
Ierent spectes. For nstence. Figure 3.27 iIluslrates
patterns 01 initial or minar deatruction o sheep
rnandtbles by wolves (in the upper two examples)
and domestc doga (in the lower two examples).
Puncture rnarks of both canine end carnassial teelh
are obvious elong the angle of the mandible. In al1
cases, carnvorea were going after the masseter mus-
ele. In Figure 3.28, where caribou mandibles are
shown. enother condtton is ndlceted. The two
lowee exemples were gnewed by dogs, bul the man-
dblea had first been strpped of meat before the dogs
receved them. The puncture maro of thelr teeth ap-
Animals may meke considerable progrese in the
destructon 01the mandible. Exactly where the man-
Another commonly citad characleristic of skulls
is the so-celled depressed fracture, which Dert
popularteed as evidence for the hominid use of tools
and the "bludgeon" technlque of hunting. I have not
personally observed such fracturas on the skulls ex-
amined from the wolf kills, but others report them te
be a regular cbaractertstc of kili and natural deeth
assemblages IHill1975; Shipman end Phillips 1976).
sufficlent of the demaged cr mscerded specimens are
available lo display .. tbe lechnique by which skulls
werespllt transveraely.Le.. by one or moreblows deliv-
ared probablywilh a scapula or Iower jaw blade on their
beses, inlo twc helves .. Each skull half thus formed
automatlcatly a CUp or bowl.
In creetures sufficienlly large such as the buffalo
these bowls. especially Ihe antartor one .. , wtth tte
horns etteched sponlaneously afforded a tripod-like
basln in whch lo mash nr trtturate any tough Iype of
FIGURE 3.27. lniliul dedruC#an af sheep by
ba!h dags !wo) (lnd wolves (uppcr twuJ_
In all these cases, there is no reason te sse the "skull
bowls' as anything other then the producl ofnormal
ettnticn by animals. It might be tnteresttng in ths
regard lo reexamine the ibex skull-horn units al-
Iegedly intentionally placed around the Neanderthal
child al TeshikTash (Movtue 1953:25, 26).
to form almost perfectly rounded, sballow. saucer
lke utenstls." Confident of the hominid workman-
shtp 00 such items, Raymond Dart (1962:287) states:
An alternativa pattern. for young animals sud
thoee with llttle lo modrate antier development (in-
cluding large animals in valvat). begins with Iba
chewing back of Iha face area toward Iha antier hase.
bul Ihere i8 also similar breakage and chewing back
from Iha occiput, leaving El "skull disc" as shown in
Figure 3.26.
Several points naed be mada abou! Ihe observad
_by canids.
The skull disc illustrated in Figure 3.26 is a comman
elemenl on many ancient sites. It i5 frequently inter-
pretad as a product al human workmanship. For in-
stanee, Breuil writes, "The frontal cavity between
Iha severed homs i5 too often carefully trimmed and
wom al Iha edges nol to haya heen utilized. This is
seen in the skull caps and human skulls of la Placard
and Laugerie Basse, snd perhaps even those at
Chou-kou-tien [1939:5j." Kitching (1963:30). wriling
sbout the fauna from Pin Hole Cave, where both
Mouslerian and Upper Paleolilhic remains are
found, commenls, "Amongsl the ten skull fragmenls
there are three frontoparietal regions of reindeer Ihal
have becn neatly removed from the tops of the skulls
FIGURE 3.28. Cranlal disc 05 clJmmonly produced by
,nawJ1I8 canJds.
... .A:..

3. Pctterns 01 Bone ModjIi;otion5 produced by Nonhumon Agenfs

Modificolions by Anolomieu! Port 65
pear dlrected at the pulp cavity below tbe molers.
The upper four mandibles were reduced by wolves
and the pallern seems to be to remove the engle o
the mandible end then rnove into the pulp cavity.
This condition has been observed by others:
Broken lower bordare of bovld mendbles need nol indi-
cote that anything has been "removed." In the natural
"non-hominid" conditions, (be lower boeders of bovid
mandibles are eventually elmcst always lost. This can
arise from the attempts of camrvores to extract the tootb
pulp, but more cornmonly tt ts the result of natural
weathering. Al leas! une deep crack appeared quite
early during the weathering procese parallel lo the hori-
zontal axis of the ramus and the developmenl of ths
reeults ultimately in a sharp break. The crack occurs
lateral to the mandibular canal. wh.ichis at the ;unction
of the basal and alveolar portions of the mandible is also
frequently lost. Breaks... across the diastemic portion
are almost universal in the carnivore-treated remains
thal I have examined [HiIl 1976:335-3361
The foregoing observation, which is consistent
with my field observations, should be contrasted
with the following statement:
lt would be absurd lo imagine Ihat the mandibles re
ceived the !realment that Ibey were given, in order to
el';lract tbe marrow, because there is not enough of tbis
malerial. if aoy. in aoy ungulate mandible lo ustify the
process. The mandibles servad in their frMh. state as
choppers. their anterior or diastllmal portions being
used as handles and Ih.ir angles as the cboppers. Afler
lheir usefulness in this capacily had ceased. Ibeir lower
hender stlllserved as pounders. their !eeth rows .. were
graten ar (Kitching 1963: 30; see also Breuil
193!'1:19-31, 62. Plate XII.
frison has iIIustrated a mandible "chopper,"
which was argued to have been used in butchering
bisoo. Its description would be basically idenlical
with Ihat given by Kitching. (See frison 1970:30,
Figure 19; also Frisan el 01. 1978:8, figures 10 and
As in the case of "skull bowls," Ihe preseoce of
"worn" edges and po1ish in Ihe absence of toolh.
punctures is nol conclusive evidence Ihal gnawing
animals did not produce the modifieations, which
are frequently cited as evidem:e for "use" on mandi-
The mandibles of Carnivorll have allmded the at-
tention of many investigalors and have promplcd
sorne of the more irnaginative interpretalions of bone
tool use appearing in the literature:
From tha similrity between the treetment of these
lower jaws of Ihe beer and len and those of the hyaena
and wolf. il ts obvtous that they were all prepsred lo
serve a common npptng purpose. Their use as tools is
confirmed by the 1111 around smoolhness of the bone
surfaces and tbetr employmenl as bcne 1001 fabricators
by the damage to sorne uf Ihe rantne teeth [Kttching
Il1ustrations in Breuil [1939:60-61. Plates IX end
X) and Ktchtng (1963: Plates 34 and 35) should be
comparad with figure 3.29 where a wolf [lower
specimeo) and an Eskimo dog mandible are tllus-
trated. Both were col1ected from wolf dens: the lower
specimen carne from the Itikmalaiyak den reported
here and the upper specimen carne from a den
visited by Eskimo informants who recovered the jaw
for me. The inlerpretation o puneture marks obvio
ously made by Carnivor canines as having been
mada by man usiog Carnivora jaws as tools seems lo
be Ihe uitimate anlhropocentrism. Such interpreta-
tions have been made on several occasions.
"'I{iURE :1.29. Wol! mondib/f1 (lowf1r fmd doS
mondihlf1 (upper specimenJ by wohes. 80111
spncimens were rccnvered !rom wnl!
vertebree heving irregular shapes are most com-
monly gnawed 00 tbe sptnes and processes. On the
atlas, for instance. the wings and the margins of the
atlantal foramen are commcnly chewed back lo their
junc1ure with the dorsal arch [figure 3.30). Not in-
frequent1y the atlas is cotlapsed intn severa] frag-
menls. Puncture marks Irom teeth seem rnost com-
rnon on vertebeee. next most cnmmon on pelvic ele-
mente. and moderately presenl on skull end toes.
Figure 3.31 i1Iustrates nicely the deslruction varteb-
rae can suffer (the upper right specmen is com-
plete). Cervical vertebrae may have transverse. spin-
ous. and articulator processes gnawed away.
Thoracic vertebras have been gnawed somewhatless
then cervical vartebrae Irom the sarna siles. Animal
marks on thorade vertebrae lake the form of broken
and gnawed tips of dorsal spines (see Upp!!r sped-
men, Figure 3.32) ando occasionally, punclures on
FIGURE3.30. AlIas ond verlebrue heavily Rnawed
by dags.
FIGURE 3.31. Lumbor verlebroe shawins Iypir;a/
of gnow;n8. (Upper right hond specimen is complele.J
Ihe body of the vertebrae. Lumbar vertebrae are not
uncommonly altered by having the transverse pro-
cesses gnawed away. the dorsal spine at!ached (see
Figure 3.31) and the articulator process gnawed or
punctured by canines.
FI{.URF. 3.32. Gnowins on dorsal spinc nd
d;stal and 01 rib.
1I !
I .

66 1. Pntterns of Bone Modifications Produced by Nonhumcn Agenls Modificalions by Anatomicol Part

g' q
FIGURE 3.33. Tooth notchins o/ rib mU'1lins by wolves.
Ribs are most commonly gnewed al the distal
ende (Figure 3.32), which may appear to beve been
broken rather then gnawed, although "mashed beck''
edges are frequent. When carnvoras disarticulate
nbe, they commonly insert their centnes belween the
ribs sud pull. This leeves notches in tbe ribs sud may
reeult in breekage generally an inch or two out from
the proximal end of the rtb: that 15, al the point of
maximel curveture as the rib leaves the vertebree.
Figure 3.33 iIlustrates rbs recovered from a single
wolf kili site: gnawing has dentlculeted tbe end of
the lower example end canina notching is well rep-
resentad on tha upper ribo
The pelvis is one area where carnivores
seemingly preferentially desteoy rnuch bone. First, lt
s an "early larget" for many cemlvores. being ODe of
Ihe firsl areas attacked duriog tha fBeding 00 a car-
CBA. Most prey animals either fall on their sides or
drop from a standing posilioo so Ibat their legs are
folded under Ihem. Feeding is therefore frequently
intialed al the "ham" froro Ihe dorsal Burfaceor, nol
uncommonly, directly inlo Ihe pelvis al the anal-
genilal area.
Look at Figure 5.04, where a recently killed
caribou is shown. It is ceer thal the teedtng began in
the anal-genital area. resulting in tbe partial desrruc-
tion of the pelvis and the subsecuent dsarttculetton
ofthe reer legs. which have been pulled beck into Ihe
picture by the pbotographer. Several places 00 Ihe
pelvis are prime targets for chewing-the dorsal
splnes of the secrum, the tltec wtngs or crest. tbe
dorsal margins of the pubic tubercle, and the ramus
and tuberosity of the tschum.
Figures 3.34 and 3.35 show the pelvis of a sheep
upon which dogs have fed. The entre carcass hed
been 1urned over to the dogs by the Navajo owners.
This pelvis iIIustrales the chorccrensttc destructioo
by a cand. Comparson should be made between ths
pelvis and those ilIuslraled from tha Glenrock Buf-
falo Iump {Frisan 1970: Figures l1i and 12J. I think
FIGURE 3.34. 1)00,01 YilJW 01 by
FIGURE 3.35. Ventral view al 5heep pely;, chewed by
veryltttle convinctng ts needed lo conc1ude thal the
pctrern 01 destrucucn is identica1. The punclure
marks so clearly visible on the sheep pelvis (Figures
3.36 and 3.37) are elso present on the bison pelves
but they are interpreted by Frisan (1970:19) as tool
marks mede by pointed choppers. Mosl commonly
the pelves al Glenrock were broken into halves. and
consisted of Ihe acelabulum with both ischium and
pubtc branches remaning E1S stubs. the lum was
veroualy destroyed. Figures 3.38 and 3.39 iIlustrale
perallel destrucncn of carbou pelves by wclves and
by dogs. The irregular denticulaled sdge where
gnawing has oecurred is also scarred wilh puncture
marks froro teeth (Figure 3,38). Analogous scar5
were inlerpreted by Frison as "tool marks from
sharp-pointed choppers" (Frisan 1970:19).
11 is nol uncornmon for pelvis elemenls lo be bro-
kenor cnl in places analogous to Ihose iIluslraled in
FIGURE 3.38. Ventral view 01 pelvis showlns toolh
punclures on the luber coxoe.
FIGURE 3.37. Dorsal vlew of Pfllvis showlns punclures
ond moshed-buck edses 01the tuber Isehlo.
! '

T ...

Modificllljons V}' Anulomjl:o/ Part 69

FIGURE 3.38. Scorlnll. plttlnll, tllld crenu/oled ediles 01
ilium prorJur:edby dog gIIow1ns.
Figure 3.39. Caribou pelvc halvee processed by man
are frequently broken across the neck o the iljurn as
well as across Ihe branches of the pubis and ischiurn
The Jatler breaks are ctean, and although fractures
and curves may occur, tha fracturad bones rarely
show the dentculeted cberecter seen 00 those
gnawed by animals (Figure 3.39). 1 think that the
pelves from the Glenrock BuffaJo lump, interpretad
as evidence o use o choppers and muscle slripping.
were simply reduced by gnawng animals.
Pelvc parts have been interpretad as choppers
(Frison 1978:307J. scrapers (Frisan 1970:19). and
mortars (Kilching 1963). 1 have reprcduced the
statement from Kilching. which becomes informa-
tve when It is kept in mind that he ts almosl cer-
tainly describing alterettons lo bones eected by the
Pleistocene hyena of Europe:
The abraded end grooved condilon of the ecetabuiar
cavtties of the 18 rhinocerns innornnale bones shows
that sorne. if notall of thern, served as mortera in which
fond was pounded and crusbed. . AH Ihe lnncminate
bones have had their lschtal and pubic portions re-
moved. and Ihe Iliac porlion remaining forros a handle,
eech of which show a higb degree of polish [Kitchng
Scapula gnawed bv animals have three basic
cllaracleristics: (a) Ihe llhe proximal tip of
Ihe spine} ISchewed away. lb} the coracoid process
and the suprt'lglenoid tuberde are generally chewed
off, snd fe) lhe edges of Ihe verlebral border will be
chewecl to varying degrees such Ihat on occasion the
blade of Ihe scspula will be all bul removed. These
three fealures are al! well ilIustrated in Figure :).40. It
should be noled Ihal Ihe lypicaJ denliculated edge
characleristic of gnawed margins is well evidenced
in Ihese spedmens. Punclure marks from animal
leelh are more common on the margins of the
glenold fossa lhan any other surface. 5uch marks
produce Ihe crenuJaled edge of Ihe scapula blade.
bul Ihe bone is so Ihin Ihal actual puncture marks are
no! genP.rally preservcd. Inslead. "circular" seg-
menfs of Ihe edlile are removed.
Fragmenls of the humerus have received more at-
tenlion by Ihose attemptng lo meke a case for bone
tools and dtstncve forms of "human" bnne break-
ege than praclically any otber bone in the enetomv.
Breuil iIlustraled distal ends of cervid and equtd
humer as having been moded by the occupants of
Choukoutien [Breuil 1939: Plate XIII). Later Darl
(1959b) was lo meke a grandor claim for "deggers"
fashioned from Ihe humen of bovtds:
From the recte presented here 1I s petent thet the an-
leJope humarus owed ils prevalence al Makapansgat
and ts persislence al Kalkbank. as well as tts dispersion
fromSouth Afrir;a lo Peking. in the Fer East and lo Por-
tugal in the Northwest. lo Ihe Iect that il was the mosl
generally serviceabla domeslic apphance discovered by
Auslrlllopilhecus [DaI11959:1121.
In addton 10 its shepe and distinctive patlem of
breekege.the fact that the proximal ertculetor end of
ths bone generally has the Ieest survival potential
and ls therefore commonly missing from bone as-
sembleges had prompted many nterpretatiuns. For
Instence. in one of the earliesl attempts lo recover
behavioral informalion from faunal rematns. Then-
dore E. whte states,
The hurnsrus shows the greatest discrepancy in Ihe
number of Ihe two ends of any of Ihe elemenls. wilh
three spedmens ofthe proximal end aod thirly rromIhe
dislal end. who has Iried IDseparal!!Ihe ScaPUIB
and humerus with a Imiteknows thal il is nol easy. el'en
in Ihase days of crucible slael, and Ihe obcan be 8&siJy
and effeclvely accomplished with a c1uaver. It is deor
Ihol o SloneAge cleover would demolish Ihe headorlhe
humerus beyond recognilion [1952:338; emphasis
Many archaeologists have accepled White's argu-
ment and inlerpreted Ihe low frequency of the prox-
imal ends as evidence of deslruclion of lhe oiot dor-
ing bulchering. "Replicalive" experimenls have
even beeo perrormed, showing thal when ane de-
slroys Ihe proximal humerus, il is deslroyed! Qlhers
have butchered animals. showing thal if one works
al it hard enough. one can even dismember an ani-
mal by puunding on arliculalions such as the prox-
imal humerus. Bul lo be sure, Ihese exerci1'ies have
, '
FIGURE 3.39. Typir:ol deslructJon of pe/l'is as produced by wolves.
3. Pnttems o{ Bonf' Modificalions Pmduced by Nonhulnan Agents
around Ihe breBkgiven 8l scalloped errad [Figure t .14b).
They were presumably used in both skinning and flash-
mg end where bone preserveton ellowed. weer in Ihe
form of 1 polish eppears un the sharp points [Frison
FIGURE 3.43. Furrowed and punctul'f!d humeral heod
removedby u wo/f while disol'dculatJns tlle Junetion with
the scopulo.
Figure 3.41 illustrates jusi how rapid the destruc-
tton of the proximal end of the humeros by animals
may be. The right front leg shown had been dragged
frorn 8" Eskimo meal cache and "eeten" by wolves in
a single aftemoon. The proximal end of the humerus
s completely gane, as s the proximal end of the
cubitus, but the fool and the lower part of the leg are
Figures 3.42 and 3.43 llustrate eharacteristic
steges of deslruclion of the proximal humeros ef-
fected by dogs and woives. Deep furrowing and ac-
lual destructon is generellv initialed on Ihe laleral
tuberosuy and the hall bead of the humerus (Figure
3:42, leftj and often resulte in Ihe breaking off ofthe
ball erttculetor surface. as shown in Figure 3.43. In
both ilIustrations, pitting and dragged puncturing
marks are visible along Ihe margins of the furrowed
areas. As gnewing continues. all Ihe 50ft lissue on
,. .
v ';,".
, ' .
, .
j \.
11' ,
, . ...A"'.
.. l-.. \
'"'-" o '<, f' lo
",., ..... ,1:
...",,--..; .-.. . .
>. _. '.,' .;,
..... ,,' " :/"i _ \
, .."
\'- . .\.'.::-...... ,, .f-";-
11>,' 1\1
J,,;,. '1'
.-- ,,-', ':.. '.....'\ -..:..."
, .. :, -r,
".",' . ..... .}t " " -,-1,: _-<
.JI- .
, , .
: ". . y , " ,
"f .. ,. "
. '''.. ,'.. . I ,-\ ..
, ' ."', ' .. \ ... '
--- f'" t' .'
"",' ,"_ ,..' I .' .'.")
t .... \ /, ,.' --
_.:, ,.
_..... .,-.
Modificolions by Anotomica1 Part
FIGURE 3.41. Caribou leg removed rom a meaf cache
and eeren by wolves.
FIGURE 3.4Z. Typical forms o/ minor destruclion (/ur.
rowing) caused by dogs on PNJXo
lmal humeri.
tool with wear striations and polish perallel lo the lon-
gitudinal axis of the bone [Frison 1910:211
The articular surfaces were removed, Ihe cancetlous
bone I!ouged Oul and the circurnference of Ihe bone
In later publtcetons this "1001" is gtven greeter
promtnence and has been largely accepted by many
early man archaeologtsts as a diagnostic item indico
ettve of human involvemenl in the formation of de-
posits and events. even though stone 10015 and othee
evtdence may be la"king. In the Cesper stte report
(1974) Frisan slates,
FIGURE3.40, Scapula gnawed by dogs (center) and by wolves-irote crenulaled edges.
not ebown Ihat men ofthe pasl ever did sueh things.
Another argumenl oered lo account for the com-
mon absence of the proximal humerus from ar-
cheeologtcel assemblages has been that it was inten-
lionally deslroyed during the production of a 1001
menufactured from Ihis bone. George Frisen
{1970:27] has ergued thallhe humerus "esher" was
a consistently produced and used 1001 on the North
American Plains:
anolher use of srnllller humerl was la break off Ihe prox-
imalarliculalor end in such a way Ihallhe edges luve a
scalloped appearance. This end was used a& B scraping
t : ! t ; . ' ~
7Z 3. Pntterns of Hum! Modifica!ons Produced by Nonhumon A,IIenls
Modificalions by J\nalomklll PlIrl
FIGURE 3.44. Humerus '1Ieshers" produced by wotves (upper) and dogs (/owerJ.
FUfllKE :1.45. hpir;(l1 (IIIilllClldf'.,frmlion nI Ihl' rndin-rubitus.
l.' I
be recalled that the articulalion belween the distal
humerus and the proximal radto-cubtus ls a rather
durable ene. in that lt comes epert relatively late in
Ihe sequence of normal diserticulation. This is re-
flected in the somewhat ahnormal survival of the
proximal end of the radto-cubttus and the typical
hreakage paUern iIlustrated (Figure 3.45). The olee-
ranon ts frequently chewed off and the proximal
radius may be punctured and pttted, the distal end is
commonly chewed off. As can be sean in Figure 3.45,
Ihe distal end frequently shears off, Ieaving an oh-
llque and pointed seclion of shaft. These may he
gnewed and chipped beck and even polshed, resulto
ing in pseudo tools of a rather convlncing formoMoat
often tbese are interpreted as daggers or plcks. Fr-
however, is the target element and is Ihen removed
from the carease: il rarely remains with tha axial
Figure 3.44 iIluslrates charactertstc "humerus
fleshers" produced by wolves and d08S. Similar
forros heve been identified from paleonlological as-
sembleges. Good examples ot humer Ihal have been
modified by nonhumen agente are gtven by MilIer
(1969: Figure 5) and Myers et al. (1980: Figure 2).
figure 3.45 iIlustrates a fairly Iypical series of
modiflcations produced by gnawing enmals. It will
animals gnawing through this jont. As we will see
in chapter 5, animals customarily gnaw through this
jnlnt on moderate to large prey. so that Ihe scapula is
normally essocteted with the axial skeleton. In con-
trast, human hutcherlng commonly results in the
removel of the entlre front quarter, and the dsertcu-
lation of the scapula-humerus joint is a secondary
dismembering operation. Even in lerge animal
bulchering, as I describe for moose in chapter 4, the
disarliculaliun of the scapule -humerus jolnt is done
flrst. because the scapula may be used for leverage
while ji ts still attechcd lo rhe carcasa. The scapule.
the proximal end may be removed until the animal
begns lo encounter the barder. denser bone of the
diaphyss or shaft, as was the case iIluslraled in Fig-
ure 3.42.
I have dtscussed tbe problem of butchering ha
scapula-humerus lotnt previously, but il should be
emphasized that oH the so-called diagnostic forms of
breekage on the proximal end o the humerus that
have been interpreted as evdence for use as chop-
pers and a "muscle stripping" techntque of butcher-
lng (compare Figure 3.42 with Figure 1.12 in Frison
(197411 are in my opinin normally produced by
y At'
J. Puttems of cne Modifications Produced by Nonhuman Agents
Modifjcotions by Anctcmrccl Port 75
son {1978:104} notes an enelcgous tem from the
Colby site---one 01 the few convincing mammoth-
butcherlog sttes in North America.
In animal-goawed assemblages. cerpels will be
eitber present in largely complete form or ebsent.
having been completely consumad. The pattern of
gnawing s lo eddress the arttculatons where the
joiot bends. If ths is done, it results in the ngestton
of the cerpels. destrucon ofthe distal rado-cubltus.
and marginal tooth scoring o the proximal end of
the metacaepel.
Metecarpals are generally chewed in two ways:
(o) Tbey may be approached as descrbed for the
carpals. When this is the case, the metacarpal te
cornmonly abandonad snce the proximal epphyas
s dense sud contene Hule grease, whereas the ex-
posed distal end of the radus contains relatively
hgh-queltty marrow. (b) Gnawing may also hegin
from the distal end: this is normally initiated by first
altacking the foot with ettendant puncturing /Figure
3.0Z} and consumpton of phalanges. The articula-
tion between the phelenges and the distal metacer-
pel is short in most unguletes. and therefore a hent
joinl is not Ihe approach of most gnawing animals.
They typically begin by chewing off the skin and
unqual phalanges and th6n work their way up to
the loes, gradually approaching the distal end of the
metacarpal. Once the cartilaginous area hetween the
phalanges has been exhausted, the chewing is gen-
eraUy stopped in favor of another bone. The
metacarpal is thus commonly found as a complete
bone with only minar damage from gnawing. This is
particulariy true at IdUs; however, in dens, lairs, and
Ihe dog yards of Anaktuvuk, less discriminaling
chewing was obvious. Such den-related destruclion
in the caribou, where the metacarpal is not a particu-
larly sturdy bone. IS through the shalt, resulting in
approximately equal numbers of proximal and distal
Fragmenta of ungulate metapodials haya sttmu,
leted the imagination of many investigators of the
Paleolithic. Such fragments have been cited as prob-
ahly used as daggers and sccops (Dart 1959a, 1959b,
1961), and they have been illustrated from Olduva
Gorge FLK North Level t (M. D. Leakey 1971: Figure
35) with abraded fractures. Use for digging or
perhaps in chopping has been suggested for enelc-
gous bones from the American Plains (see Fr-isen
1970). As in the majority of other tool cletms. the
evidence consists of suggestive breakege and. in the
case of the Olduvat material, abrasion and rounding
of the edges, which is nterpreted as use. As we have
seen repeatedly, these are properties that rnay be de-
veloped in a faunal assemblage without the partici-
pation of mano
Figures 3.46 and 3.47 iIIustrate several femoral
fragmente Ircm wolf kille and from Eskimo dog
yerds: Figures 3.46 and 3.49 iIIustrate sheep femora
gnawed by Navajo dogs after they were fed an entre
cercees during the winter of 1971. Severa! charac-
leristics that are very regular are illustrated: almoet
without excepnon the greater trochanter is chewed
off the proximal end by the gnawing animals (see
FIGURE 3.46. Distal femora fram wolf IdUs
icol pom:turins onri furrowil1i1of thc lrochleo.
FIGURE 3.47. Distal femur showinB tooth punctures
aJong medial sur/ace o/ Ihe di,,'al condyle.
FIGURE 3.48. Proximal femora showlns wnsis-
lenl remoyol of thc sreater trochonter by dORso
FIGURE 3.49, DJ.stal ylew o/ prodmal ldIeep femur
showillS tooth punctures and /urrnwl1!S'
Figure 3.49). Animals may then prcceed to chew
through the neck of the femoral beed end f1nally fur-
row out the entire proximal articuletor end. Figure
3.46 illustrates the distal ende of Iemora removed
from wolf kills end these without excepton (that ts,
of damaged bones) exhibit the furrowed removel of
the medial ridge of the trochlea. (Compare this pe-
ture with Figure 8.6 in Frison 11978:3111.1 Both
characteristics have been cted as evtdence for mus-
ele stripping as a human butcherlng procedure. Por
instance, the following "reconstrucucn" of a hutch-
ering procedure, largely based 00 the ideas ofGeorge
Frisan, is given by lohnson:
frison poslulaled that processlns of the hind lag besan
with loosenins of the palella and removal of Ihe tubar
calcls of Ihe calcaneum and the trochanler majar of the
femur. This may also have been the pattern on the Llano
Estacado as this Iype of damage was common, wilh the
exception Ihal Ihe luber calcis was generaly not re
moved. Damage lo Ihe ca1caneum (distal projeclion
cracked or broken off) appeared to be from lIttempts lo
sever Ihe lateral and middle annular ligaments. Afler
this WllS accomplished. the ca1r;aneum would then be
free to serve as a handle wilh which to strip vBrious
muscles [Johnson 197a:1031.
It should be noted thal all the "diagnostic" pat
leros af destruclion cited as evidence for musele
are commoo in my canid-gnawed as-

76 3. Pcttems uf Bone Modificafions Produ;eo by Nunnumon Agents
Brellkage UnroJoleo lo ConsumpUon: Trompling and Bone Manipulalion
FIGURE 3.51. Typical odvallced riestructjuII 01 th!.' proximal tibia showinS challnaling nnd chippinS bnck.
By ths time the reeder can probebly anttctpete
the types of destructton SUD for the various bones.
stnce the pattem is redcndant: biased destructon of
the soft bone parte. particularly the artculator ends
of long bones. The tibia is no excepton end ths is
well illustrated in Figures 3.50 and 3.51. The tibial
rrest is generally the Iirst projecton lo exhibit gnaw-
ing, end snce it is such a prominenl projection it is
frequently removed by aligning the bone parallel lo
the tooth row end vstng clown with the cemesstal
teeth. This produces B semicrenulation when viewed
Irom the sirle, as is the case for both spectmens on
the right side al Figure 3.50. The left specimen
shown in Figure 3.50 rllustrstes a reletively ed-
FIGURF. 3.50. Differenl views o/ animal on
'he proximal libia.
vanced state of destruction-the en tire proximal end
Is removed. 11 should be noted that the proximal
ends of the tlbial shaft are chtpped back elong the
chaoneled edges. producing a "denticulated" and
"retouched" series of edges (Figure 3.511. Typically
this level of deslrucHon precedes any majar altera-
tions 00 Ihe distal end. which is comrnonly 10ft ar-
uculeted wlth Ihe tarsals. Ir destruction conttnues.
the shaft may be chenneted back and collapsed so as
lo leave the distal end with a mini mal segment of
shaft attached (note such patterning in Tabla 4.07).
When such an advanced stage of destructton is
reached. there is an almost invariable pattern of
chlpping beck coupled with poltshng. as shown in
Figures 3.10 and 3.12, occurring en segmente of Iho
broken ends of the attached shaft. Very similar forms
are described by Frisan as tibia choppers (see Fig-
ures 1.17 through 1.19 in Frison (1974)), Given my
knowledge of animal pattems of bolle destructton. 1
belteve thal il would be unusual for an animal lo
attack the shaft of the libia before effectlng sorne
modiflcation on the proximal end. This observaran
renders it more Itkely that Frison's tibia choppers
manufactured Irom the proximal libia are not animal
produced. Nevertbeless. the tibia choppers tltus-
trated in Figures 1.18 end 1.19 (Frisan 19741 would
certainly be suspect in an assemblege wheee there
was other evdence of animal gnewtng. Examples of
"tibia choppers" given by Slanford (1979a, Figures 6
and 7) are textbook examples of enimal-modfied
The most common form of deslruction seen on
larsals is Ihe gnawing away of Ihe distal end of Ihe
calcaneus [tuber calcis). Puncture marks are occo,
sionally seen on alher tarsols but by far Ihe mosl
common Iraces of animal gnawing are on the (;al-
Pallcrns of meta!arsal deslruclion are verv similar
to thosc noted for Ihe metacarpal. lo Ihe d08 yard
sampJe fmm Anakluvuk. Ihere were numbfJTS of
melalarsals that wera channeled almost Ihe complele
length of the bone. with the channeling having
started al Ihe proximal end [the lower four spec-
mens in Figure 3.52). 1have not seen 5uch systemati.
cally channeled specimens fram either fhe wolf dens
al Ihe kills (see Figure 3.171. I suspect tha( the dogs
gnowed the bones in this fashion largely in response
lo having been fed already disarticulaled lower legs.
because Ihe ooly area not covered by skin is the prox-
imal end, and I have observed the dogs preferentially
slar( ealing al Ihe exposed end. The paltern of con-
sumption described for the melacarpa\ is belleved lo
bemore common among "wild" animals consuminS
Iheir own prey. Neverlheless. Darl (1967;120/ illus-
Irales sorne nearly identical specimens from a
Mousterian assemblaRe recovered in Israel. Figure
3.53 shows the distal ends of melatarsaIs recovered
from wolf dens and the channeling is obvious. The
specrnen at tbe lower righl is channeled and there
are punclure ffiarks and pilting on the opposite side.
As pointed out for the metacarpals, Ihese areclassic
eXBmples of "scoops" as iIlustrated by Dart (1959B,
1959b, 1961).
Breakage UnrelatOO to Consumption:
Trampling and Bone Manipulation
Thus far 1 have discussed bone breakage as if lt
occurred primarily during consumption and carcass
~ . , ..
. &
3. Pctterns of Bone Modifico!ions Produced by Nonhumcn Agents
BreekegeUnre/ated lo Consumption: Tromplingand Bone MonipuJotion
FIGURE 3.!SZ. E"treme channelins of metatarso" by lethel'f!d dtJgs.
dismemberment by animals. OC course Ibis is not tbe
case, and the aeoheeolcgtet must be continuously
Bware o the possibility Ihat alher agents and con
lexts of breakage may candtion what he OI she ob-
servad. C. K. Brain (1967bj was one o tha firsl to
poiot to tha important role that trampling might play
in botb breakage and the production o pseuclotools.
Diane Gifford {t977) has made important observa.-
tions on tha role o lrampling by large animals. par-
ticuJarly fOI bones depositad along tha margins o
bodies o waler. This is o considerable importance
when behaviorlll c1aims such as those macle by L. S.
B. Leekey (1968) ere constdered. In describing his
materials from Fort Teman, Leekey stetee,
Ihere Bre In Iha &lime deposit, and at the same level,
small areas oC fossils where the bones have been broken
up, and where Ihe damase indudes excelltll'll examples
of depressed fractures of Ihe types usually associated
with "a blunt instrumenl" 11968:5261
Leakey then goes on lO say,
The available evidence preSllnled there, Iherefore.
slrongly suggests thal lhe llpper Miocene hominid
FIGURE 3,53.-"'ol{ deslrut;/ion on distol melolorso/s,
commonly citad pseudotoo/s.
Kenyapilhecus wickeri was alreedy maklng use of
'Iones to break open animal ,kulls in order lo get al Ihe
breln and bones to get at Ihe marruw 11968:5201.
AH the breakege descrbed by Leekey is assan-
tally of the same type observed by Diana Gifford as
derived from the trampling of exposed bone by liv-
ing animals. The interesfed researcher should com-
pare Leakay's photographs, particularly Figure 3,
with lhose presented by Gifford (1977: Figures 68-
Prom both the Old World (Biberson and Aguirre
1965) and the New World (Bonnichsen 1978; Irving
Bnd Haringlon 1973), there have been daims that
"worked" or "modified" elephant bone represents
tools manufactured or used by man. A recent "ex-
perimental" butchertng of an elephent was con-
ducted by severa! persone interested in bone tools
teee Park 1978) as a demonstration that tools manu-
factured from elephanl bone could be used in hutch-
erng an elephant. This teern o expenmenters. led
by Dennis Stanford of the Smithsonian staff con-
cIuded that an e\ephanl could indeed be butchered
with tools manufectueed from ts own bones (Parir:
Such cletms rnust be weghed ageinst reseerch
findings where actual controlled conditions have
been sought: thst Is. where the condtons of forma-
tion or production of traces or remens are known
end one seeks to recogntee d agnostic characteristics
of those conditioos. Such controlled research has
been provtded by two sourcae (Cae 1978, 1980:
Dnugles-Hamiltnn and Douglas-Hamilton 1975) re-
gerdtng elephents and the possible condilions of
breekege and modiftcation dtrectly referable to the
taphonomic conditions of elephant death, dsertcu-
lation, and burial. O particular importance is the
work of Malcolm Cae (197B), who actually took the
pans to observe what happens to en elephant car-
cass after death. He observad in detail the progresa of
decey end tha scettertng of the rematas of elephents
that ded natural deaths. The deeth sttes of elephents
are under certetn condtons sttes of violent ecttvtty.
Coe observed that "the large petch of disturbed
ground around most corpses observed (diameter) in-
dtcated that death was usually protracted after col-
lapse [1978:77J." Falling and protrected thrashtng
(Coe 1978:77) by an elephent s cartainly a condition
that could result in green-bone fractures.
The deeth sttee of elephants are apparently Ire-
quently visitad by living elephents. who tremple ex-
posad bone and disperse such bones by knocking
them ebcut with ther tusks and even kckng them.
Cae (1978:761 observed crushing o rib bones from
trampling, presumably by elephants, approximately
ayear after the control animal's death. Another in-
teresting comment by Cae is as follows:
The sitelelalfemsin9 still persist 00 aHsites 9tudied flve
years after the allmsl's dealh, Long bones and situlls
have been scallererl up to 100m. flOm the d,olh 9He by
alher elephofl!Swhile smaller bones have been chewed
nlo smal1 frllgments by hyoenlls.... Siteletons Iying in
! 1
i I
j ',r:'
- r
J. Potterns of M'odificolions Prnduced by Nonhllman Allenls
Ihe cpen and subjected lo greet diurnal temperature el(-
tremes are severely fssured and flaked tCoe 1976:79;
emphasis mine l.
In arder lo gan sorne appreciation for he mag-
nttude of dtememberment and scattering that can
occur al an elephent death sne. the inlerested reader
is raferred lo the remarkable cclecuon of photo-
graphs of e\ephants in various states of decay and
dispersin pubhshed by Beard (1979).
Of particular importanee lo those interpreting
breekage of elephanl bcne ts the mysterious e"cile-
ment elephants exhib! over the bones of deed
alephanta. The following description illustrates Ihe
behavior of elephants toward lhe bones of their dead:
They aH began Iheir delailed olfaclory examinaBan.
Sorne pieces were TOcked gently ta IInd fro with Ihe
farefee!. Dlhers were knocked togelher wilh a wooden
dank. The lusks exclted immediate Interest; Ihey were
picked up, mauthed, and passed [rom elephanl lo
elephen!. Orle immalure male Iifled the heavy pelvis in
hi, lrunk 9.nd cllITied it forfifly yards befaredropping it.
Anolher slllffed \wo l"ibs inlo his moulh and revolved
them dowly as if he were tasting the surface with his
longue. The skull was roHed over by one elephant arlel"
anolher. To begin wilh on\y the lal"gest individuals
collld gel near Ihe skeleton. ,ueh was Ihe erush.
Boadeceala Ilamedelephantl arriving late pushed lo Ihe
cenler, pickPd up olle of the lusks, Iwiddled il for a
minute or so. Ihen carried it away. wilh Ihe blunt end in
her moulh. The rest oflhe group nowfollowed. mllnyof
Ihero earrying pieees o{ the sKeleton which were al!
dropped wilhin aboul a hundred yards lOouglas-
Hamillon and Douglas-Hamillan 1975:2391
This hehavior, coupled with direcl trampling as
noted in the studies by Gifford (1977), could cer-
tainly he expecled IDgenerate bone breakage. De-
panding on the candition of Ihe bones heing Iram
plad Of manipulaled, green-bone or spiral fraclures
might be regularly expected.
Particularly germane to Ihis point is Ihe pre-
viously ciled work by Myers et 01. (1980).
f. J' P f..pl Q j L eA 8S well as
pollshed, chipped, and otherwise modified bone in
deposits where"nnm had,c,ei tsMj plsjld I'RJtMI!F1n
Of more interesl.
however, is their observation ,hal evidence for
predator-scavenger relaled modificalions was rela-
tively infrequent. Thev suggest tbet tramplrng R'Nl.y
well nave been a major cause of the reletively 18r8e
nurnber of spiral fraclures and other mod#ications.
They experimented with a wealhered bone: "We
have found that slighlly weathered hnnes break quite
easily bul sttll exhihit the charactertstc spiral break-
age pattem of resh "green" bone [Myers el (JI.
These researchers Illustrete Gonvincing spiral
fractures produced by trampllng the bones of a cow
tha had been weathered under natural conditions
and approached weathering steges designaled No. 1
or early No. 2 in Behrensmeyer's (1978) study of
bone weathenng. This research opens up a malar
pos!'>ible source of bone modifications unrelated lo
human activ'ity Ihat has receivl'd very Hule researeh
altention. This may be particularly important to a
number of c1aims regarding the human involvement
in Ihe modificalion of elephant bone. Biberson and
Aguirre (1965) attempled lo duplicate pallerns of
breakagc among elephanl bones recovered from the
important site of Tort8lba. Bonnichsen's (1978,
1979) analysis of alleged tools from the Old Crow
Flals locality places slrong emphasis on modified
elephanl bone. He cites a number of elepnant bone
f1akes Ihat have aIl the properties of having beeo
removed by percussion techniques analogous to
lithic reduclion strategies. Similar forms have been
reported by Slanford (1979a, Figures 8, 12). Noted as
companions lo these finds are a number of elephant
bone "cores" from which f1akes have beco removed
(Bonnichsen 1979:101-129). Morlan (1981), after re-
studying the faunal remains (rom old Crow Flats.
hes basically accepted the argumenl that Ihe
modifiu!otions of bones (including spiral fractures)
(rom animals smaJler than elephants may well be
referable lo the actions of predator-scavengers. He
has dismissed Ihe role of trampling and bone man-
ipulation by elephanl!'> as nol likely 10 produce
"poinl loading," that is. localized impacl scars. He
conc1uded thal Ihe carnivores Ilre loo small essen-
tialJy lo address effedively eiephant bone, and Ihat
(ractured elephant bones. particularly specimens
exhibiling "poi ni loading." are the product f
human modificalion and there(ore tools. No ,.riti-
ci!'>m of this argumenl is vet available. Neverlheless,
the dismissal of and bone manipulatio
remains en opinion totally unsupported by empirical
sone Modificotions ond Methodology
material: thus Morlan's argument s not a vald ar-
gument from eltrntneton, a point 1 will tale up
somewhat laler.
Bone Modificatioos end Methodology
Throughout Cbepter 1 1illustreted case after case
of magtnatve inlerprelaton COf observattons made
en Ihe archaeologtcel record. Must of the lime m}'
aim was to demonslrate a hidden assumplion-
namely. that man was the egent responstbte for the
particular modifications or associalions being "in-
terpreted" in behavloral lerms. Both Ihe premise as
lo the identity of Ihe agenl and the interpretalions
offered as to the condilions in Ihe past responsible
fur the characler of the Ihings observed in the ar-
chaeological record were, in Chapler 2, shown lo he
inferencfls. In Chapter 2 it was also argued thal aU
our ubservational language and the means available
lo us Cor in(erring lhe past from observalions on the
archaeological record had to be rooted in experience:
There are two ways of regarding concepts. both of
which. are necessary fo Ilnderstanding. The first is that
of logical analysis. Il answers fhe queslion. howdo con-
cepls and depend on eaeh ol"er? !n answer-
ing it we are Dn comparalively safe ground. But this
seeurity is pllrchllsedal lhe price of emptiness of con-
lenl. Concepls can only acquire contenl when !hey are
connected. however indirnr.lly. wilh sensible ex-
peril'llee.. no logleal invesligallon can rl'\'eal lhis
conner.lion. l can only be experienr,ed [Einslein
It was argued that middle-range research with Ihe
aim of developing t1'JJiable melhods for inferring (he
:tasi had, Iherefore. fo be raoled in aclualislic
This chapler has reviewed certain claims Ihat
others have made regarding Ihe meaning lo be at-
teched lo properties o( modified bone. Calegorically
speaking, lhc daims cenlered on Ihe degrce lo which
lhere are dislinclivc forms oC bone bmakilge referable
exclusively lu Ihe aclions a( mano aml on Ihe degree
lo ..... hich pallerned modificalions observed on vari-
ous bone could be reterred lo mndificatinns pro-
duced by num in orrler lo tuiC thl" halles as lools.
The advocates of "osteodontokeratc" (Dart 1957),
"protoltthtc" (Menghin 1931), "pre-paleol thic'
[Ncolaescu-Plopsor and Ntcolaescu-Plcpsor (1963),
a dsttncttve bone-using culture "Alpinas
Palaolithlkum" {Bachler 1940). or more recently an
early and continuously irnpurtant bone technology
in the New World (Irving and Harington 1973) gen-
erally imagine bone modified by man as having
many aatures in common with ltthtc technology
(Bonnichsen 1978). Ths analogy has served archee-
ology poorly for severa! reasons. It s true thal many
of the modifications observad in bone are the result of
fracture dynemcs. or the mecbentcs of impact leed-
Ings or pressure loedngs. as are the basc dynamics
of lithic fracture. Both materials m8Y be studied in
terms oC Ihe same general questions regarding the
physics o( fraclure. However, in nature there is a
fundamental difference: Lilhies are apparenlly regu-
larly used by only mano 11 is true that lilhies may
become (ractured in a variely of inanimate, mechani-
cal conlexls in nature: pounding of stones on
beaches, in friclion "cauldrons" in rivers. as a result
oC sluffing off and falling ofrocks from cave ceiling8,
cliff (aces, and Ihe m.e. In aH these conlexls the
stones being subjected to fracturing forces are being
modified as forees act on objecls and the orientation
of the objects is generally free lo V8ryindependently
of the direclions of impacl forces. lf nol, the coinci-
dence o orienting forces and (racture is relalively
random. producing rare and irregular
of fracture palteros Ihal stalistically form a compo-
nenl of a more comprehensive pattero a( random
breakage configuralion. This is in marked contrast lo
slones modi(jed by man, who orients the stone ..nd
directs the impacl so thal a regular pattem of orienta-
lion Cor both the piece being acted upon and the
direclion of aclion or impact loadiog is maintained,
resulting in distinctive and palterned modificalions
in the slone being acled upon. Such'patterning gen-
erally permits the discrimination betwfW!n tools pro-
duced by man and slones modified in nature. Trans
ferring this opprooch lo bones cloes not work, For
unlike slones, bones are regularly manipulated
mechanically by animals olher lhan mano The resull
is a paltero of regular. mechallically produced
modificalion in bone Ihat shares Ihe configuration of
redundanl or luading relative lo redun-
danl orienlaliull. in chipped edges. pieees
3. Puttema af Bone Modificalians Produced by Nonhuman Agenls 82
that have multiple flakes originating at a common
mpect eurrece, mltiple ekee removed in redun-
dan! ortenteton. both with regard to the dtrectton of
impact and tbe orienlation of impacting to tha prior
shape of tbe pece. ln Iact. ell the patterntng contigo
urations of tools are seen. These are trua enetogues.
in that tbe formal configurations have many things
in common with stone tools but the causes may be
quila diferent. No! seriously conetderng ths. many
have argued that they are homologues; that ts. roan
was using his documentad knowleclge of mechanics
on both slene and bone, er in sorne cases only on
bone, and producing artifacts. Arguments frOID anal
ogy ossume Ihe causes oC analogous characteristics
are Iha same. This is nol !roe for fracture patterns on
bone versus stone.
Failing lo w(;ognize the strong analogi<:al com-
ponent to resear(;h reasoning, sorne researchers have
thougbt tha( they could gain a greater undershmding
o tbe lIignificance o modified bone by conducting
experiments with the mechanics o bone fracture. A
numbar o detailed and "controlled" studies have
been aimed at increasing our understanding of the
properties of bone that (;ondilion ls fraclure poten-
lial when mechanically manipulated .(soo particu-
larly Bonnichsen 1973, 1979). Although these are in-
teresting and the knowledge gained can be used,
lhey in no way infonn us about Ihe causes, in the
proximal sense, of the loading agents responsible for
bone breakage. Nor do they provide us with any
criteria for recognizing such agents. Yat these were
ocluolistic sludes. They were "experimental ar-
chaeaiagy" of a relatively elabarate farm. How cauld
Ihey nal provide us with methods? The answer, af
course, is lhat too overall procedure one uses does
not ensure suecess. One must ask productive queso
tions and one must succeed n gaining adequale con-
trol5 over experimental observations lo be able lo
relale Ihem accurately to "causes."
Making irrelevant observations, in terms of either
the accurale isalation of causes or characteristics nol
observable in Ihe archaeologieal record, will not con-
tribute toward Ihe development of usable ar-
chaeological methodology. Ths is a major problem
many so-called ethnoarchaeological sludie.
They are simply malerially orienlen elhnographies
and do nol attempt seriously eilher lo isolale causes
or demonstrate Ihe neees,>ary relationships between
"bears" of tnterest and "toctprnts'' actually recover-
able from the archaeological record.
My approech has been very consciously aimed al
not falling into Ihe trap of belevlng that f a study is
actualsttc. it must be useful. I sought lo obtetn con-
trol ccllecttons of bones known lo have been rnodt-
fied by anmals and to describe and ldenufy the ef-
fects when the causal egents were known. 1 have
used the spectmens whose properties could be un-
ambiguously referred lo animal manipulation to
demonstrata the ambiguily of rnany tnerences made
by others.
1 think 1 have demonstrated thal spiral fractures
are nol unique lo man; Ihat spiral fracture when pro-
duced by anmals is not limited lo an origin at the
distal ends af the bones; that modifications ciled by
Frisan at Glenrock as evidence of muscle stripping
are commonly produced by gnawing canids: and
Ihat the characteristic modifieations commonly ciled
for recognizing "expediency tools" are regularly
produced by nonhominid predalor-scavengers; and
Ihat, finaI1y, many of Ihe modfications cited as evi-
dence of human modification and 1001 use are refer
able to predator-scavengers.
My interests go beyond the evaluation of specific
ideas to archaeologieal melhodology in general. Are
there guidelines for research Ihal, if more explicitly
underslood, could reduce Ihe numbers of mvths
generaled by archaeologists? 1think Ihis mus! an-
5wered affirmatively since archaeologisls have. in
fact, had very HUleexperience in Ihe research area of
methodological development. For the mosl part,
Iraditional archaeology did nol recognize midclle-
range research as a doma in distinet from general re-
search. Methods grew up largely as a series of ae
cepled conventions for giving meaning lo the aro
chaeological record. Methods "accumulated" out of
Ihe mosl common research situation faced by Ihe ar-
chaeologisl, in which he recognizes patterning in the
archaeological-paleontologir;al record and must faee
the nlriguing problem of whal the pattern means.
For on!y with a solution to Ihis problem can the ar-
chaeologist converl contemporary panero observa-
tions into meaningful slatemenls aboullhe past.
Most often this problem is faeed by Ihe ar-
chaeologisl after he or she has oeseribed Ihe patlern-
ing (frequenlly in a seelian labeled "IYIlOlogy" or
"slatislical analyss"l. Once palterns of association,

Bone Modifica!ion5 and Melhodolagy
coveeeuon, and so en heve been establsbed. the ar-
chaeologtst may then wrtte an interpretattve sectton.
This is where the ercheeologtst may build a model of
the past or offer opntons as lo what condtons in the
pest "ceused'' the propertles summerized. This s
what 1 refer to as a post 'roe cccommoduve argu-
mento Surnmarizing Ihen, this is a situation whera
one has isolated petterntng (Ihe effeets) and seeks lo
use imagination to envson the condrttons that. if
Ihey had beppened. would account for Ihe propertles
seen [the causes). The way one werrants Ihis con-
struction of Ihe past may be eomplicaled and draw
on considerable knowledge, even theory, from other
fields. Nevertheless, lhe degree of fil between he
magined condilions and the obseMled properties in
Ihe archaeological record is nol a test of the oeen-
raey of lhe argumenl. Mosl perSons are knowledge-
able enough to invenl a siluation that, if true, would
account for the observed facls.
In order lo convinee others, a number of warranl.
ing arguments are frequently proposed. Sorne are de.
signed lo demonstrale thal the suggestion is plausi-
ble and not totally unrealistic. This is Ihe role in
which ethnographic analogy is cast by Thompson
(1956:329): "He must test his conclusion by demon.
slraling Ihat an artifact-behavior correlation similar
lo the suggested one is a common occurrence in
ethnographic reality." However, the degree of
plausibi/ity is no! sufficient to establish probobility.
Probability s frequently argued in Ihe form of an
argumenl from eJimination.
Argumenl5 from eliminalion have two basic
premises: (o) AHthe polenlial causes are known and
Iisled and lb) al! but one of those lisled are not the
cause of the phenomenon in question. Therefore, the
remaining cause is considered lo be Ihe correct one.
lf one could not juslify Ihe premise lhal all Ihe possi-
ble causes were lisled, Ihen Ihe conclusion Ihat the
cause nol eliminaled !hrouRh rejeclive argumenl is
Ihe correel one conld not he sustained. For instance,
1 mighl argue !hal Ihe causes of a particularly in-
leresling pile of rocks could be (a) that it had been
marle by spacemen, (b) that il had been r:onslruded
byead, ann(e] lhal il had been prodllced by geolog-
ical aclion. [ rilen miRhl proceed lo eliminale alterna-
ttves a and b through warrented argument. end
Iherefore conclude that the pile of rocks was pro-
duced by geologtcel aelion. I would be very apt tu be
wrong: snce there s no reason to belteve thet all of
the potential causes hed been Iisted in the first place.
Of course, in arder to construct such a Iist we would
have lo understand causaton, am f we had such
understendtng we could certainly devise a more
powerful end Informativa eveluetve procedure.
Argumenl from elirnination wes the form impl ed
by Bonnichsen's rejeclion of the alternative Ihat the
spiral fractures observed at Old Crow Flats had been
caused by animals breaking their leg5. His is a very
common way of warranling an opinion regarding a
preferred post hoe accommodative argument. Nor-
mally one uses imagination and an accumulated
knowledge lo think ol all !he possible situations Ihal
might have accounted for Ihe observations made.
One Ihen offers an opinion as to which is considered
mosl likely by denigrating all but the preferred ar-
gumen!. This seems to be wha! many researchers
mean by Ihe "method of multiple working hypoth-
eses." This is a fonn of warranting argument for
one's opinions, but il is in no way sufficient for pro-
viding us with methods of n(erenee.
1must add Ihat for an argument from elimination
to be valid, one musl have available unambiguous
means for monitoring the alleged "causes" and
therefore a way of aetually detennining the degroo of
parlicipalion by a suggested cause in a system of pasl
delerminaey. In lraditional archaeology Ihe argu-
menl from eliminalion was used as a form of war-
ranting argument for one's beliefs about the pas!. In
Ihe new archaeology, it should function more to war-
rant one's melhodological research: "A valid propo
sition can only be refused Ihrough hypolhesis test
ing. However, when faeed with valid alternatives,
one can evaluale in prohabilislic terms Ihe relative
slrenglh of alternatives and make decisions as to
how lo invesl research lime IL. R. Binford
A particularly interesting form of argument from
elimination is one where only one possibility can be
imagined. Mosl Ihis argumenl lakes Ihe
form Ihal "only man could do thar." It is generally

84 3. Penems of Bone Modifications Produeed by NonhumanAgents

Bone Modifieotions rmd Methodology
executed by citlng complex or regular patterned
mcdtcettons in nature. which seern self-evident in
their manfestatlon of design or planning: "1suppose
that these bcne breeks could come about wthout
human ntervenuon.... But Ihe broken mammoth
leg bones we've found-nothing could do thet ex-
cap! humane hurling rocks {quole from Denns Sta-
lord published in Canby 1979:3541." Commenting
on thls fOfID uf aegument. two distinguished proles-
sors of logc stete.
Argument trom want of evtdent aitemaves Is, how-
ever. ene of the most abuslld argumen! forms. Often
when we "jump lo we are ahusing il; W6
leap al Ihe nrsl Ilxplanatory hypothesis tha! COffil!S to
mind without duly surveying Ihe field IQuint> llnd UI
Iian 1978:120-1211.
This form of argumenl is so easily abused because
only one alternative is recognized. The only positive
form of argument is offered with regard lo allerna
tives, which are rejected, so Ihere is no direct evi-
dence Ihal can be offered in support of (he conclu-
sion. This is the form of argument used by Von Oani-
ken (1969) to c1aim extraterrestrial involvement in
lhe history of culture. Quine and Ullian (1978:121)
comment germanely on this poi ni: "Indeed argu-
ments of Ibis kind (argumenl from want of evident
alternalives) are a favorite device of charlatans, if not
absoluleiy indispensable lo Ihem."
I think Ihal il should be clear fhal, although Ihese
forms of argumenl may appear impressive aud even
convincing lo many, unless Ihey are aclually exe-
culed with means ror measuring Ibe alternalive
causes or actually evaluating the relalive conlribu-
Iioos of differing variables said lo have been active in
a "conditioning" sense, the enlire exercise rernaios a
form of polemics and is in no way an epislomologi-
cal procedure. It slands as a ser of ranked opinions,
not a stalemenl o probable determinacy.
The techniques frequenlly employed in demon-
slralinglhe plausibility of a post hor: accommodative
argumenl are usually suspecl. As menlioned earlier,
Thompson suggesled Ihe citation of ethnographic
precedent. Whal if there is no elhnographic cilaticln
relevant lo Ihe paHerning you observe? This 15 par-
licularly important when very early
eras of hominid expf'rience. It is Isocommon wilh
respect to the types of pattems erohaeologtsts my
observe. Such things rarely prompt commenls Irom
ethnogrepbers and historians. We have a fne exam-
pie of Ihis situalion in the work of George Frisan al
Glenrock (1970). Frisan observed a large number of
petteened modifcations in bone. There were no de-
tailed descriptions of such Ieatures in the etbnog-
raphc record. He mede the assumption Ihal man was
the causal egent. He Ihen invented a set of conditions
that. f they had happened in the past. were beleved
lo accommodate all the "facts" of bone modtceton.
Uslng his exceptonel knowledge of animal anatomy
and manual tool use, Frisen tmegmed a proceder
far butchering using "expedient bone tools." largely
choppers and f1eshers, Cor muscle stripping {he ani-
mals. The patterns of bone modificalion noted al
Glenrock could Ihen be underslood as resulling Crom
chopping loose muscle attachmenls on bones, as well
a5 wear patterns and intenlional modifications made
in shaping and using bones as lools in Ihis operalion.
Frison Ihen engaged in aclualislic
licalive experiments. That is, Frison (1974) actually
bukhered a bison using Ihe proceduces he had imag-
ined and concluded Ihal his model was wel1 con-
ceived in {hal the properties expecled lo result Crom
his imagined operalions did in Cad resull. Olhers
Ihen adopled his model oC butchering as a methodo-
logical device for giving meaning lo spedfied prop-
erties of Ihe acchaeological record. O{her researchers
began finding expedienl tools and evidence or mus-
de stripping. A melhodology was being molded by
lhe adoption of a 5el of conventions Cor giving mean
ins lo properties of the archaeological record based
on Ihe Cailhplaced in an individual's work. The rep-
lieativa experiment showed that Frison's knowledge
and judgmenl used in developing his model oC Ihe
pesl was sound. It showed Ihal his model of Ihe pasl
was plausible. JI in no way demonslraled Ihat it was
As poinled out earlier, Oart condueted ceplicative
experimenls together wilh Kitching. concluding Ihal
Ihe uack-and Iwist melhod of breaking bone was
responsible for spiral fractures. and he was wrong
Sadek-Kooros (1972) enndueted replicative experi-
menls ano concluded thal man intenlionally broke
bones lo a paltero for use as lools, and she was mosl
cerlainly wrong. George Frison conf!ucled replicat-
ve fJxperiments wilh rcgard In his ideas abOlll mus-
ele stripping as a mode ofbulchering and ccncluded
that in Iect he could muscle slrip a bson and smu-
late the pattems of bone breakage and destruction
noted in prehlstoric assemblages. 1 argue that he is
most lkely wrong. Recently there have been severa!
"replicative" elephant buteherings (see Huckell
1979; Park 1978). One of these was led by Oennis
stenfcrd, who is very interested in illustrating that
an elephanl eould be butchered with 1001.'1 manufac-
lured from its own bones (Park 1978:94).
Under the epislomology being ..
ive experlments provide an evaiuation of the [udg-
ment of the person who proposed the original model
oC the pes!. If one can produce the effects by fol1ow-
ing a procedure proposed in explanation oC cerlain
observed properlies, Ihen Ihe proponenl's knowl-
and skill in Ihinking are demonslrated. We
could Ihen place sorne confidence in Ihe argument
because oC Ihe demonslraled skill oC Ihe advocale in
anticipating Ihe consequences oC certain operalions.
The demonslration of plausibilily, eilher Ihrough
Ihe cilalion oCelhnos;aphic procedent or through
rl'plicative experirnenl eoupled wilh warranling ar.
guments by elimination as a linked batlery of laclics,
wasused by many archaeologisls lo convinee people
lhal their views of the past were correcl. These tac-
lies primarily serve lo demonslrate competenee:
The final judgmenl of any Ilrchaeologist's cultural re.
conslruclion musl theterorebe basadon ltnappraisal of
his professional competence.and pllrticularly lh!! qual-
ily of Ihe subjectivf! conlribulion lo Ihal compelence.
Our presenl ml!lhod of assessingIhe role oi lhis subjec.
tiveelemenl by an apprlllsal of lhe inlelleclual honesty
of the arehaeologisl who milites (he inferenees is cer-
11Iinly inadequllte. Bu!. lhere dot's nol seem to be any
praetil:al means of greatly improving the situa_
lion We can only hope for improvemenls in !he
mathodsuf measurinA Ihe amount of failh we place in
an work IThompson 1951"'131-332].
Isu,:sesled long ago lL. R. Binford 1968b: 16-17)
halthere musl be ways olhar than Ihose adopted by
lraditional archaeologists for evaluating archaeolog_
ical argumen!. Herf'! I add Ihe suggeslion Ihal Ihere
musl be research enrleavors designad sPf!cifically far
the devc10pmenl o infeffJnlilll me!hods. and Ihese
IAnnot be easily eVHluatedusin,Ilrchlleological dala
nar cal! they be adopled fmm "respected"
archaeologtsts. This rneans thal tnducttve erguments
developed post hoc lo account for petternlng ob-
served in Ihe ercheeologcal record must be Irealed
as provoeative ideas in need of evcfucuon. Only
after they have been eveluated can they be ceunously
adopted as observanonel lenguage. or nterpretatve
conventions, or methods for giving meaning to the
I suggest that middle-range reseerch progresses
by virtue of the eccumulauon of knowledge !hal in
faet permita one to use an argument by elimination
properly. Jf actualistic research con be cited sbow-
ing thal wha! is ob.'lerved in Ihe archaeoJogical re-
cord is not refemble to a suggesled cause. then one
bosis for myth making will be eliminated. We must
conlinuously Cocusour allenlion on the properties 01
Ihe archaeological record and, in turn, COCUs our iro-
aginalions on those observed propertiep>. Once we
have generaled a post hoc accornmodalive argument,
our research task is specified. How can we investi-
gale Ihe idea advanced as lo Ihe context 01causalion
for specified properlies oC !he archaeological record?
Where in Ihe COnlel1lporary world or the domain oC
recorded dynamics can I seek experiance facilitating
Ihe modeling oC causes Cor a particular pattem ob.
servable in the archaeologieal record? This is Ihe
firsl Question. The nexl question one must ask is Ihis:
Even iC one can solale such a souree oC enHghlen-
menl, how ambiguous are Ihe properties for which
one is seeking a causal understanding? Thet s, are
aH Ihe possible causes lisled? This is a necessary
condition (or an argumenl from elimination. Not
only musf Ihe researcher face Ihe possibility of the
ambiguily oC Ihe pallerns as observed in Ihe ar-
chaeological record, bul he or she must also faee the
equally crucial question o Ihe relevance o actualis-
tic understanding lo Ihe pest. Can we make uni-
formitarian assumplions from our contemporary
underslanding regarding Ihe production oC prop-
erles thal appear to be common lo Ihe pasl and Ihe
present? These are aU complicaled problems tha!
must be faced in order lo develop a slrong methodol-
ogy aorl ereale an unambiguous observational lan-
guage for Ihe science of archaeology. Failure lo arl-
dress Ihem when making melhodological c1aims will
genprally conlribulc to Ihe building oC modero
myths ralher Ihan lo Ihe gcowfh of archaeologicaJ
Pllr111. MiddJe...f!onge Heseurch-e-ln Senrch oIMelhodoJogy
Interpretnons were offered and judged "probable" or plausible they fre-
quently become conventions whereby additionaJ observations nt new
sltes were lnterpreted. Grodually o myth wcs built up about the post. The
rnyth conststed of inferences drown from unevaluated premises ond Its
very scale o/ ccceptcrtce graduolly became further justification for beJief
in the myth. Unfortunately this remoins one o/ our dominant
"methodologies. "
In Chopter 2 the point wos made that we necd to de
mefhodologies cnd observational lungucge through mlddle-mnge re-
secrch. This reseurch was envisioned as Iorgely uctucllsuc. where con-
troJJed information about causes and ejfects could be evaluoted experien-
tclly rather than inferentialJy, as has been Ihe common pructlce. Method-
olcgicc! resenrch requires that both the becr ond the footprint be observa-
ble. Methodological resecrch in service of crchoeclogy must normalJy be
conducted with living systems [ethnocrchceoiogyl or enes in which the
refevcnt dynamics hove been recorded (historicol orchceclogy] or where
the relevanl dyncmlcs may be replicoted (experimental orchaeology).
The polnt of view cdopted here Is thct octuolislic studfes or middle-ronge
reseorch is crucial to orchoeology and should be conducted from the
perspective of the archoeological record. We ore not studying material
"residues" (R. A. Gould 1980:250-251) or "behavioraJ correlates" from
the perspective of the ongoing living system. This view contrasts mork
edly wilh Ihat of Schiffer (1976), Rothje (1979), and StiJes (1977), who
seek o science of material culture, or reJationships between behavior and
material culture. 1am not suggesting that studying contemporory syslems
from a materialist's perspective should not be done, oniy thal doing so is
nol archaealagy; it is ethnogrophy and as sueh faces very different meth-
od%gical and intellectuoJ problems. Stated another woy, we are not
aUempting to specify the relotionships between "behavior" in any
exhoustive sense and material remoins. Instead, we are attempting to
understond lhe of patterning and various struclural prop-
erties of the archaeoJogical record in arder lo leorn about lheir post.
Jt was suggested in Chapter 1 Ihot we need methods for assessing the
inlegrity (the number of identily of Ihe ogents tha! aeted in the post to
produce the deposit being invesfigoted) ond the resoJution of a deposit
(lhe number of and redundancy in Ihe events represented by the depo-
sits). Tbe bulk ef this book is concemed'tN1tfnhedevelopmfmi.t>f'11t9tftods
for identifying and the
predator""Scavenger' poltrottthft'-'ontt"J"atetmtotogiunl
depoS'its. This is basicolJy o prohlem in identificotion.
Fol/owing the suggestions developed in Chapter 2 l wiIJ describe ac-
quired properties of bones and patterns of assemblagf! variability that are
referable lo known agen!s. l will atlp.mpt to justfy Ihol sorne properties
are uniquelv referoble to specific agents, and will sUMest sorne of the
Prlrf JI. Heseurcb-c-ln Senrch ofMcthodology
behavioral characterislics of the ngents that ensure the uniqueness of the
prcperes descnbed.
The majority of materiaIs described in the following three chepters
were collected during a 4-year study of the use of onimals and anima}
produets by the Nunamiut Eskimo of north central Aloska. (See L. R.
Binford 1976, 19780, 1978b, 1979, 19BO; L. R. Binford ond}. B. Bertrom
1977; and L R. Binford ond N.}. Chcsko 1976 for reports on this reseorch.]
The focus of the Nunnmiut study was the relationship between thetr hunt-
ing ond consumer stmtegies as execured in funclionoJly and seasanally
differing sites (L. R. Binford 197Bb) and the faunaJ remains al those ses.
The study wcs middle range in choracter, aimed at developmg methods
for recognizing sites of vorying funcuon from faunol remains.
Chapter 3 summorizes a considerable body of cbservcons on bones
modified by animaJs. As indicoled in Chopter 1, mony controversial
cloims regarding the nalure of the POS! hove been inferences from obser-
vations on broken and modified bone. 1 wi/J revlew sorne af the "inter-
pretative Jiterature" where faunal remains have been central to crgu-
menls abouI the post. This material is presenred because the contents of
Chopter 3 ore germone lo (hese orguments, bu! olso because my knowl-
edge ofthese arguments has guided much of my ohservation on bone. As
Dorwin is credifed (Cohen and Nagel, 1934: 197) wilh having said; "How
odd if is that anyone should not see that aH observation musl be for or
ogoinsl sorne view, if il is fo be of any service." Chopter 3 treats o number
of the properties of bone modificotian about which a vos, number of
inferences regarding the post hove been prompted. Since 'here hm'e beeo
so many claims, much of this chopter will appeor negotive. J will s}'s
temoticoJly cite properties of bane breakage coupled wilh surficiaJ
modifications produced by either dogs or wolves ond compare lhem to
specimens or condifions thal others hove cited as evidence [or humon-
hominid modification or manufacture. Although it ip; quite true that much
afthis chapter appears la be concerned wilh discrediting the inferences
made by others, 1 am simultoneously presenting descriptive malerial in
sufficient delaiJ lo provide a bosis for lhe identificarion af lhe agenls
responsible for modificalions thal may be of inleresl in future research.
In a very real sense 1 hove alreody written a book on the topic covered
in Chopter4--what men do with bones. As you might imagine, 1wiJI draw
heoviJy on Ihis previously pubJished moterial, but sorne new facts will be
presented. My orientafon is one of comparing human behavior wilh ani-
mal behavior on the one hond, and evaluoting claims a!her or-
chaeoJogisls hove made regarding Ihe meaning to be attached lo certain
formol potterning in ossemblag:es of modified bone on lhe other.
Chapter 5 is concerned with rhe central probJem of the book, the. djag-
nostic.differences betw.een banas modified by man ond !hose modifie,d by
onimols. 1will present sorne original ohservalions on wolf hehovior and
"I'; "'!
1' 11
, .
'1' 1,
,:, 1
'1'. i
1: 1
11,' '1
i i'
I ' ,
1 .... r,
,; J"

3. Paems of Bone Modificalions Produced by Nonhumen Agents
The task Di giving meanng to observetlcns s very
dlfficult. We saw in Chapter 1 how "control" was
obtained by inferring the identity o a causal agent
Irom an aseocraton with tdentteble tools. That as-
surnption provided so intellectual anchor. Pattern-
ing observed in assodated thngs was then reerred
to the behavtor of the rnferred common agent. Bones
beceme Iood, trequenctee of speces became stale-
mente of dtetery preferences or hunting strategies.
Patterned rrmdificatcns en bones became tools.
In this chapter we haya seen how "control" was
getned by ustng argumenl from analogy for forms of
patteming observed erncng stone tools and bones
and assuming thet analogoue petternlng was refer-
able lo Ihe same causal conduons. That was wrong.
We S8W how replcatve experiments and nedequate
ettempts to tnvesugate alternative suggesttons
coupled with elmtnattons by "opinion" served to
justify tbe use of conventtcns. such as spiral fracture
equals man, ar expediency tools equal rnen. or de-
strcved trocbenters and proximal humeri equal mus-
ele strtpplng. We must expiare ways of increasing
our eccuracy end reliability in making inferences to
the past. Tradtional erchaeology did not reccgntze
the need for, nor even the possibilily of developng.
middle-range research. Traditional approaches are
demonslrably tnadequete.
It seems lo me that the method ot rnost rapid pro-
gress ... s going lo be lo sel down explicitly al each
step jusi whal the queson te. and what all the alterne-
Uves are. and then lo set up crucial axpertments to Iry lo
dtsprove sorne.... When wbole groups of us begin lo
concntratelike thet I belteve we may, see... tncreases
in the rete of scientific understanding [Platt 1964:3521.
Since ths chepter was written I have learned of an
argument startlng in Australia regarding Ihe agents
responsble for abone assemblage unaccompanied
by stone tools (Archer. Crawf14rd, & Merri lees 1980).
Assuming thet one can ettrbute lo the hand of
man all remains in a deposit essocated with ac-
cepted artifacts was shown lo be a common method-
ologieal error in my revew of literatura presented in
Chepter 1. Incorrectly identifying the agent respen-
sible for modifications on bones was a problem em-
phasized in Chapter 3. In this chapter I will demon-
strale that Ihe assumption of monocauselity has been
a common error associated with the interpretation of
bonee modified by mano Here the ercheeologst ob-
serves properties of the archaeologtcel record. for in-
stance modfed bone. and assumes that all the
modtceuons arose from a single behavior. such as
butchering, merrow cracking, cr tool producron.
In most cases we may expect that bones and other
archaeological remains had alife history during
which they were componenls of a cultural system.
Differing aclions were taken wilh respect lo Ihe ma-
terials and in turo they were altered differenlially,
depending on Ihe characler of Iheir Iife histories. We
must view an item, fealure, or sile as being the ac-
cumulative resull of numbers of differenl acUons and
events such thal the overall morphology of Ihe mate-
rial remains carries information about many dif-
ferent modifying conditions-not a single causal
evenl bu! a sequence of causal processes. With re-
spect lo bone Ihis view has nol generally charac-
terized Ihe analysis of archaeological remains.
Chapter 4
Human modes of
bone modification
! 1
. ~
4. Human Modes o/ Bone Modifkolion
PrevicusApproochesID UnderslondingHrnken und Modified Bone 89
Aside from thts ontologtcal point regardtng ar-
oheeclogtcal materials, 1 will continua with rny
methodological lnterests. 1 will explcate entena for
recognizing the agents responsibJe for materials that
might be Iound etther in association with unques-
ttoned tools or remetns o hominids themselves, or
independently. Tha! te. although my goa1sare meth-
odological they are also fecused 011 the geoer!c df-
ferences between men sud ober animals in Iheir pat.
terns of modifying bone. For this reason. in the mate-
rials lo be presentad I will rnake no attempt lo dts-
cuss tbe designs or the tecbnques actual1y used by
man in Iha production oftools from bones. 1am sol e-
Iy concerned with morlifications unidad 00 bone
as a by-product oflhe explaitatian of animal foods by
mano In one sense J am only inleresled in Ihose ac-
lions man perforros in solving proMems Ihal musl
also be solved by olher animals in Ihe utilization and
consumplion of prey species. I am furlher limiting
myself lo a concern wilh Ihe lechniques lIsed in dis-
membermeot and meat removal, and the lechniques
used in breakin!! bones for the recovery of bone mar-
Within Ihis dornain 1will address the problem of
maldng uniforrnitarian assumplions about Ihe mate
rial derivativas of human behavior thal may be ob-
served actualistically. What is the role of ethnoar-
chaeology and ethnographic analogies in middle-
range research?
With respecllo the specific subjects covered, the
available infarmalion is by no means balanced. Fol-
lowin!! a survey of previous researeh and a discus-
sion of Ihe dala 1will be appealing lo in my descrip-
tive discussions of human use of bone in a food con-
texl, I will begin Ihe descriptive seclions wilh a dis
cussion of dismembering stralegles. This is, regeet-
lably, a very shorl section despite much discussion
in the lilerature ofbulchering lechniques. 1wiH Ihen
lum my atlenlion to cut marks on bones, whicn are
more commonly referred lo as butehering marks.
This is a long, delailed seclion summarizing Ihe de-
scriplive lilerature as known to me and presenting new
material. It will be painfully c1ear Ihat HUle i[lforma-
tion is available on this suhjecl I will add lo Ihe
available malerial by presenting the grouped dala
from Ihe Mouslerian levels al Gambe.,Grenal cxca-
valed by Franc;ois Bordes (see Bordes 1972). as well
as the data from lhe Kakinya site exc8vated durinK
my work with the Nunemiut Eskimo (see L. R. Bln-
ford 1978b:374). The relavely few ceses of de-
scnbed materels are drawn from a remarkably wide
geographical and temporal range: Iwo Mouslerian
cases. a case from eastem Nnrth America during tbe
Early Historie periodo a Mississippian site from Mis-
souri, several Plains btson kili sttes. a Plans an-
telope klll ste. and my materials from the Bskimo.
Over al! Ihis material 1will be able lo damonstrate a
kind of remarkeble redundancy in the plaeemenl of
cut marks and their particular forms. Thus, 1 will
venture sorne empirical generalizations about
human patlerns of butchertng end the telltale marks
remaining from Ihe use of cutling 1001s in aceom-
plishing this lask. "The plausibility of a hypothesis
depends largely on how compatible Ihe hypothesis
is wilh out being observers placed at random in Ihe
world IQuine and Ullian 1978:74J." The diverse na-
ture of Ihe samples gives sorne plausibilily lo Ihe
generalizations offered.
Finally. I will discuss marrow cracking as il has
been described by others and as I have witnessed and
documenled jI among Ihe Nunamiut Eskimo. Much
of my discussion will center not so much on the
production of generalizations but on the degree lo
which premalure generalizations by others have mis-
led analysis in the past. I wil! make a pies for Ihe
colleclion of more informalion on Ihis subjoct and
offer sorne suggestions regarding formal atlributes
!hal might wel1 prove useful in fulure melhodologi-
cal developments for dislinguishing the hand of man
from Ibe jaws of animals.
Previous Approaches to Underslanding
Broken and Modified Bone
Reviewing earlier research on human modifica-
tions in bone shows thal sorne of Ihe earlies! re-
search was on cul marks and surfidal modifications
resultiog from man's use of 1001s in butchering
(Martin 1906:10). This eMly worlc was nol, to my
knowledge, followed up un'i) relalively recenlly
(Guilday n.d.; Guilday el nI. 1962: Parmalee 19(5).
Mosl of lhis work is sound and provides case
.'iludes of hulche'ing marks.
Most of the reseerch orienled toward the "recen-
structton" of prehistoric butehering practice stems
directly from the seminal work ofTheodore E. White
(l9S2. 1953a, 1953b, 1954, 1955). White suggested a
number of post hoc accommodetve modele lo ex-
plain the differenlial frequencles of anatomical parts
revealed tbrough the eareful anatomical identifica-
tion of bones recovered from archaeological sitas.
These models took Iwo baste forms, postuletmg {o)
Ihe differenlial trensport or abandonmenl of anatom-
leal perts in the context of bunting logistics, and [b]
the destructron of anatomice! parls durlng the
bUlchering process, wbich was beavily modl'lled on
the assumplion of regular use of large choppers and
heavy cleaver-like lools and bafted mauls. Butcher
ing practiees Were generaIly considered lo vary elh-
nically [see Wood 1962) and to be a eharacleristic
thal could be used for Iradng the culture history of
differenl elhnic groups.
Typical of tbe post hoc modeling of Ihe buteher-
ing proeedure is the following statemenl from one of
White's early papers 00 Ihis subject:
The humerus ShOW5 the greatest discrepllncy in the
number of the 'wo ends of any of lhe elements ... Any-
one who htls triad lo separate lhe sCtlpula and humerus
wilh B knlfe knows thal it is not easy. even in lhese
dllYs of crucible slee!. IIlld lhe oh ClIn be ellsily and
effectively ar.complished wilh a cleavllr. It is clear Ihal
a Stone Age deaver would demolish lhe head of Ihe
humerus heyood recognition. further culting up of Ihe
forelimb appears to havebeeo lIccomplished by smash
ing the radius and metacarpal near the middie as the
ends are usuaUyenlire [White1952:338}.
Here we see the post hoe accommodation Df a
butcbering model lo the patterns of breakage and
bone deslruction observed in an archaeological as-
semblage. It is assumed Ibal tbe bone destruction
resulted Crom butebering activities. White's posl hoc
argumenls were general1y adopled. For a number of
years reading an archaeological site report that in-
eluded a summary of faunal remains was Iike read
ing a litany. Wbile's posl hoe argumenls were in-
voked lo "explain Ihe arcbaeoJogical facts" and
Ihereby inlerpret Ihe pas\. 11 ls interesling in Ihs
regard Ihat Thomas Kehoe reporled Ihal his nfor-
manis as well as historieal aceolmls ndieale: "The
limbs were cut up al Ibe joinls, nol Ihrough Ihe
bones ... there was culting through the scapula-
humeros joinl /Kehoe 1967:69-711."
As menfioned earter, there is running through
the Plans literatura a very strong "normattve" bias.
Such dtscrepenctes as tbe one Kehoe reportad,
elhnohislorical information indicative of a prnce-
dure eonlradictory lo the post hoc model. are fre-
quently trealed as evidence for "cultural dif-
terences" between the peoples responsible for the
Iwo sourees of Informeuon. There hes been little al-
tempt lo determine experimentally. ethnoar-
chaeologtcettv. or from historie records f there are
any siluational correlales of alternative bulehering
slralegles (see L. R. Binford, 1978b:87-90. Little at-
tenlion bas been given lo delermining if there are
diagnostic properties that might permit Ihe ar-
chaeologists lo diseriminale differing slralegies, or if
Ihe arehaeological faels far which posl hoe interpre-
lalions bave been offered did in fael derive Croro
butchering activilies!
I will relurn for a moment to one of my inilial
poinls, namely Ibal Ihe bones recovered from asile
may hava undergone a number of differing event
sequences during which modificalions could have
been made prior lo Iheir having entered Ihe ar-
chaeologieal record. In the last chapte.r 1argued fhat
the bison bones excavaled by George Frison al the
Glenrock Buffalo Jump had becn modified by
scavenging animals, presumably after the site was
abandoned by mano In that siluation sorne of the
propertif'!s of the bone probably derived frOI1l Ihe
bulchering activities of humans and sorne {rom tbe
scavenging 8clivities of Ihe animals. lf we recognize
that bones may be differenlially acfed upon by men
in various contexts, and thal the same acts con
ducled in different conlexts may be carried out in
different ways, the "interpretative" picture becomes
even more unclear.
EtRR98n:1. eh8iel' .,._h (l.. R. Binford
1978b; Yellen 1977a) thal provided new information
on butchering practices general!y also included in-
formalion on mOre Ihan just butchering. Indeed Ihe
studies also included deseriplions of "animal pro
cessing" such thal the life histories of bones EUld
their accumulation of properties were empirically i1.
lustrated. This knowledge has not yet prompled re-
search leading fo eefinemenl of methods for analyz-
ing faunal remains relative lo butchering practices,

11, e,:f.J.J.. el" \, , \
{ V .' .
Jj_ '. 'j.,
-:U f ,1" \ )
Befare dtscussmg the potenel of such reseerch as 1
see it, il ts necessary lo revtew studes amed al
understanding properes cf bones derved from
rnerrow cracking sud other p-ocessng steps such as
bone greese manufacture, bone uice producton. and
o bone nto con..umable formo
I haya already mentioned the replicatve expen-
menta by Dart and Kitchng (Dart 1959) as they relate
to the "proof" of argumenta in favor oC a distinctive
. mode cf human bone breekage, the ceeck-end-twst
method. lt is hard lo believe that there eppeer to haya
heen no ethnographic accounts of bone breakage by
man available to reseerchers al that time. To rny
jnowledge, Ihe first epecic study aimed al under-
standing marrow-bone breakage was the study men-
tioned earlier by Zerhut (1967) and participeted in
by (1973). In terro' of published ec-
counts the work o loho Yellen [1977a) among the
Dobe !Kung provided anothar eccount of bone
breakege, but no deteiled descriptive informetion re-
garding the morphology of the derivative fragments.
Little analytical attention was given to discrminat-
ing fragments produced during different processing
phases of !Kung bone use. My own re,earchers
among the Nunamiut (t978b) presented sorne de-
'criptions of marrow-bone breakage but 1did not ap-
preciate al the time the naed to obtain conbollad
samples for purposes of recognizing properties that
derived from one phase of bone use snd breakage
versus another. 1was interested in assemblage com-
posilion snd the differenlial use and survivorship of
hone elements in differenl contexts where bones
were used and manipulated. That is, 1 viewed my
research tesk as Ihe identification of the functon of a
site from the overall configuration of liD assemblage.
II WlJS nol until laler that I began lo worry about
morphological properties of broken bone as clues ID
specific behaviors.
One experimenlal study (Sadek-Kooros 1972,
1975), one replicalive exercise (Bonnichsen 1979),
and two patlern-recognition studies of erchaeologi-
cel bone (Lyrnan 1978; Noe_Nygaard 1977) basically
complete Ihe Iist of research available lo the studenl
of patterned bone breakage.
What emerges from this brief survey is thal with
the exception o research on "butchering practices"
there has in Eacl beeo very Hule argument regarding
the meaRing to be atteched to pallerns ofbone break-
4. Human Morles of Bone Modificalion
ege- As will be pointed out in the various parta of
ths chapter treattng different bone-modifying ac-
tions. these properues have generallv been viewed
monoceusallv, We need methods for discriminating
dterent bone-modifying situations one from
enotber. many sucb situatrons may well be reflected
morphologically in a single bone assemblage.
The Control Data
The new material s lo be presented in this chapler
suffer from having not been collected with the spe-
cific questiolls addressed in this chapter in mind.
With only minor exceptions lhe ethnoarchaeological
observations were all made belween 1969 and 1972
while I was carrying out fieldwork among the
Nunamiut Eskimo.
As part of lhe Nunamiut study much lime was
invesled in observing butchering and lhe dismem
berment of caribou and moose especially but
mountain sheep and bear as well. In all, the hones
remaining on the sites of over 400 butchering epi-
sodes were recorded. Detailed step-by-step observa-
tions of butchering procedures were made on 37
separateacts ofbutchering animals conducled by the
Nunamiut Eskimo. I personally field butchered 13
animals under Eskimo supervision using differenl
tools and strategies so I could leam the problems
associated with different taclics. Much of this work
has been presenled (L. R. Binford 1978b) and will
only be discussed here in relation to animal behevior'
or in the contexl of illustraling variable petterns of
bone breakage associaled with different butchering
make use of several well-documented Es-
kimo bone assemblages to ilIustrate relative frequen-
des of cut marks, and as comparative dala relative to
animal-produced assemblages. AIl these as-
semblages have been previou!lly described and re-
ported and Ihe reader will be referred to the appro
priate pages in my earlier book for the behavioral
Importanl lo my Eskimo research was a knowl-
edge of the techniques and procedures used in break-
iog marrow bones and processing bone for other
__ _ --I!f;-'
purposes. sucb as rendertng bone grease. At the time
the observations were made from which data will be
drewn. [ was not thinkng of the marrow cracking in
terms of comparisons wilh bones broken by animals
nor was 1 rhtnktng in terms of animals otber than
canbou. which was the terget speces of my research
among the Nunamiut. That ts. I was trying 10
generalize about cerfbou, not aboul the variety of
animals relevant to interpreting the remans al Dld
Crow Flals or Olduvet Corge or other interesting
Pleretocene sites. My marrow-cracking investiga-
tions were designed to provide a baseline for
evaluating whelher or no! arttculetor ends from
caribou long bones had buen either destroyed or re-
moved from sites. I was concl."rnedwith establishing
an estmate for the expecled number of splinters and
chips produced by bone-breaking Eskimos per
long-bone articulator end. Thus a series of marrow-
cracking evenls were conducted during which en
Eskimo woman and aman broke a series of long
bones using lhe repertoire of techniques normally
employed by the Nunamiut. The aim was to establish
sorne estimate of the differences between splinter to
arliculator end ratios for different bones of the
caribou anatomy. AHsplintcrs, chips, and articulator
ends were saved and catalogued from two such con-
trolled situalions. In four other available samples aH
lhe breaks were produced during marrow process-
ing. but I do not know the exact anatomical compos-
tion of Ihe populalion ol bones broken. What is lack-
iog in alJ these samples is control over the actual
breaking tactics and tlJe resulting bone splinters. 1
was asking population questions regarding the
properties of aggregale assemblages in my fieldwork.
Whal we would Iike to know is the characler of aoy
diagnostic morphologicaJ properties referable to dif-
ferent methods of bone breakage. Unfortunately, we
do nol know Ihese Ihings.
Dismembering Sfralegy
Man using tools is capable of actually taking an
animal apart lo suit his purposes. That is. he is not
dependent on the nalural slrengths of the articula-
liaos and the amounts of connective tissue lo sepa
rete parts Ihat then may be differenliaJly used or con-
sumed, as in Ihe case with predetor-scavengers. This
is not to say that the amounts of muscIe and the
character of connecttve tissue do not condition
men's selecton of perts or the tacttcs he used Icr
dismemberment: they most certainly do. The point ts
thet the particular anatomicel sets that man manipu-
lates are conditioned by how he dismembers a car-
case. which in tum is condtttoned by its intended
Several workers have recorded the way dlfferent
efhnic groups perttuon the anatomy o animals, par-
ticularly ungulates. Diane Gifford (1977) recorded
the basc units nf the anatomy into which four sepa-
rate African groups partitioned Ihe anatorny and 1
have recorded actual hutchering data for the Navajo
and the Nunamiut Eskimo (L. R. Binford 1978b; L. R.
Binford and J. B. Bertram 1977) lohn Vallen (1977a,
1977b) has observed butchering and hence the
characler of segments generated by Bushmen. Figure
4.U1 compares the ways in which these groups parti-
tion the anatorny of animals during butchering
Several poinls are of interesl;
1. AH groups ideally separate the head from Ihe
neck between lhe occipital condyle and the
atlas vertebra.
2. AH groups except Ihe Navajo separate the nack
from the remainder of the vertabrae
3. AHgroups separate the front leg from the axial
skelelon. Those that furlher parlition the front
leg into upper and lower segments generaJly
disarticulate between Ihe carpals and the distal
4. Al! groups separate tbe rear leg from the ver-
lebrae: however, there is considerable var-
iability in the degree to which half Ihe pelvis is
left attached to the rear leg as opposed to the
axial skeleton. In all lhe ethnographic cases
recorded where either axes. adzes, Of large
c1eaver-lype knives were used in bulchering,
the pelvis anrlJor sacrum were bUlchered off
with lhe rear leg. Where small knives were
used, the pelvis andlor sacrum were butchered
off with the lumbar vertebrae or as a separale
lInit distind from the rear leg.
5. Al! groups generally treat the spinal column
wilh ribs and brisket distinct from the other





FIGURE 4.01. Dismemberment pror:rJretl by vorious elhnlc groups durlng butcherillll.
majar poruons of the anatorny, but there is
considerable variability in the way the thorex
and spine section is butchered clown into
smeller unts. Most. bu! not all, butcher the
rihs end sternum off as independent untts.
Sorne of the vartabtlty tlluetreted in Figure 4.01
is ltkely to be relatad to degrees of cognitiva
speclficity-that s. the leve! of dscrtmneton et
which nforments perceved that the question was
directed. In the case of observed butchering, there s
a knd of normative "Ievellng." in that under sorne
condtone the Nunamiut and the Navajo dscrm-
nated down to elements even ner than those indio
catad for the Kalinjmo. Taking this problem into
consideralion, there is a fair degree of corre
Djsmembering Strategy
nenls such as house. meat rack. end dump; onIy site Ieted in MNIs dvtded by the maximum MNls ob-
totals were utlteed. Tbtrty-one variables represento
servad for 80Y variable in each assemblage. This pro-
ing the anatomical parts ltsted in Table 4.01 were cedure ensured that there would be no variance con-
used in the anelysie. Values used were percentages tnbuted to the malrix by differences in popuJalion
of maximum MNI, or the ratio of the variable tabu- size among the cases. A BMDfactor enalvtc program
TARLE 4.01
Factor Loadin,s for AnoJy&fso/ Nunamiut FounaJ Assembla,es
Anatomical part 1 2 3

5 e
1 AntIer -.13 02 -.11 .09 -BS -.08 .23
2 Skullimaxilla -,12 -.03 .Q1 -.02 -., -.10
:> Mandtble -.39 -.04 -.19 .07 -.63 .38 -.11
o( Atlas/axisvertebree D9 -.40 .22 -.16 -.42 -.40 -.27
S Cervical vertebras .2' -.26 30 .00 -.51 -.36 -.15
5 Thoracic vertebras .94 .Q1 .07 -,03 D5 .05 .01
7 Lumbar vertebsee .91 -.05 19 -.oa 09 -.02 .02
8 Sacrum BS .02 06 -.19 02 .02 .23
9 Pelvis .81 -.12 .15 - .02 05 .30 -.03
10 Sternum

.05 -,28 -.12 21 -.32 .19

11 Ribs .72 .31 -.08 -.22 .10 -.10 06
12 Scapula .39 -.21 .41 -.09 .14 .30 -.19
13 Proximal humerus .35 -.31 .73 .12 .03 -.05 .21
14 DIst81 humerus .13 .11 .87 -.07 -.01 -.01 --.03
15 Proximal radte-enbttus .03 .08 .85 -.10 .07 -.09 -.14
15 Distal redlo-cubtus -.04 33 .63 .08 .0' -.30 .2'
17 Carpals -.16 .08 .28 .44 .11 .08 .52
18 Proximal rnetacarpels -.18 .14 .41 .18 -.14 -.66 .05
19 Distal metacarpals -.07 .24 .24 .37 -.22 -.60 .04
20 Proxlmel fmur .12 .8' .22 .05 .00 -.11 -.12
21 Distal Iemur .07 .83 17 .16 .01 .05 -.08
22 Prolllrnal tibia .0' .78 21 .17 .13 -.06 -.12
23 Distal tibi81 -.21 .'9 -.02 -.10 -.04 -.01 -.36
24 Tarsal! .09 .15 .12 .42 .02 .53 -.41
25 Astragalus -.26 .39 -.00 23 19 .17 -.56
26 Celceneos -.20 .29 .10 25 .21 .14 -.75
27 Proximal metetersat -.17 .50 -.oa 11 -,04 -.14 -.51
28 Distalmetateraal -.15 .3' -.05 .41 .09 -.12 -.63
29 First phalanga -.13 .08 -.04 .89 -.07 .09 -.06
30 Seconrl phalange -,12 .05 -.05 .94 -.01 -.03 -,09
31 Third phalange -.06 .08 .01 .90 -.01 -.17 -.11
Cornmunality 825 82.7 746 66.7 652 88.9
Thorax Upper Upper
Head Metacarpal Tersels
and reer front and and end and
splne leg leg podials "OCk upper metatllrsal
NOTE: Factor dillMnostics shuwn in itali!.;.
spondence among the ethnic groups in the elements
intc which the anatomy is partitioned. Although
provocatve. these data are either normativa or besed
on informant information and as such are suspect as
gutdes to actual variobility that might be generated
under specific condmons.
To obtain a cleerer pcture of the elements of the
anatomy that might be expected to be treated as units
and as such perhaps to have tndependent histories
under conditions of human uttltaaton, a factor
enalysts W8S carred out on 64 unedited assemblages
of anatomical parts col1ected among the Nunamiut
Eskimo. These were basically all the assembleges re-
portad in Binford {l978bj minus dog yards. When 1
sav the assernblages are unedtted. thls rneans that
assemblages were nol broken down into subcompo
4. Human Modes of Bone Modifjcalion
94 4. Human Modes of Bone Modifiootion Dismembeting Stmtegy
fiGURE 4.02. Johnny Rulland.dripp;ng lendon bundle Imm me,apodial.
FIGURE 4.03. Johnny Rulland holding strlpped unt: IOnJals and 1001,
independent dstrtbutons of the metatersels and the
metacarpals 8S well as Iheir independence from Ihe
upper leg segments of both the rear and forelimbs
clearly reflectthe independenl processing of marrow
as opposed lo meet. and the differential quality of
front versus rear quarter "marrow" durtng periods of
poor nulrilion among the cenbou.
1do nol offer the Nunamiut data as an example of
human palterning only lo tllustrete Ihat human use
of anatomical elemente is apt lo heve certain redun-
dant pettems. to evidence sorne spectalzed usege. as
in the differences between marrow- and meal
yielding parts. and to be segmentally grouped in
lerms ofuse consideralion. The conlrasl must he kept
in mind wilh processes of"natural" dismemberment
as documenled by Hill {1976:151). In the contrasts
rest diagnostk criteria.
interesting petterns of organized or parallel "deter-
minance" are the grouping of phalanges with Ihe
carpals end tarsals (factor 4). This pettern derives
from charactertstcs of Nunamiut processng of
bones for marrnw (see L. R. Binford 1978b:148j. The
foot s removed by servering the altachmenls on the
dorsal fece of the metapodial between the articule-
tlon of its distal end with the phalanges. The Ioot s
then held in one hand end tha distal end of Ihe
melapodial te g:lfSp'd)in the other. The sheath of
tendons nmntng in ff{e ventral channel of the can-
non bone is strtpped up to Ihe uf Ihe
hone. Cuts are then made between Ihe
bone. resulting in Ihe removal of the or tar
sals aHached lo Ihe tendon sheath, which in luro is
atlflched lo Ihe phalanges. The enfire unil is Ihen
disposed of. (See Figures 4.02 and 4.lJ3.) Finally, the
documented by Hill (1975). Of particular interest is
Ihe fact that the thorex and vertebrae group behave
logelher (factor 1) with the exceptton of the cervical
vertebme. The letter are demonstreted lo intgrate
wilh tbe skull and mandible (factor 5). This pattern
of covanatlcn ls most unlikely among assemblages
where "neturel" processes of disarticulation domt-
nete. Reference lo Table 3.01 illustrates that the
cranium with alias vertebra is disarticulated from the
remander of Ihe neck quite early (step 6) in the se-
quence of decreestng organizalion, and the neck
separares Irom Ihe thorectc vertebras al ebout the
same time Ihe ribs begin lo seperete tstep 16). This
means Ihal under natural condnons elernents IhRI
disarficulate al aboul tbe seme time are mosl apl lo
end up essoceted end such a paltern s in strong
contrest with that noled in the Nunamiut dala. Other
was employed using orthogonal rotatton option. The
results as indicated by factor lcadngs are sum-
marizad in Table 4.0l.
There is a seven-factor solution with a cleer
eigenvalue "jump" between the seventh factor
(eigenvalue 1.56) sud the next factor (eigenvalue
0.97). The cumulative percentage of variance BC-
counted for by the seven Iectors is 75.2%. The com-
mualities for each factor are given in Table 4.01. This
solutlon appeers very clean and s technically a
robust pattem-recogntton etetement.
This solution is provocativa sud rnost informa-
tiva, given what is known ebout Nunamiut treatment
of animals. OC even greater nterest are the dif-
ferencee between the anetomtcal segments Ihal be-
lb I.J haya independently in the Nunamiut data and the
i/ .., natural diearticulation sequence (see Tsble 3.011
4. Human Modes of Bone Modificafion Hutchertng Mnrks 97
TABLE 4.02
Djlllribulion o/ CUI Marks Oll Bones in Five Nunomlul Assembloges"-"
'MN'[ m;nlmal nllmher of elaman!1 in each calagory: MKO = numb!l1' mllTkedby cut marh
'Valullll far pelvis, lK:apuJe.an<1 varlebraa mal' well be deOalad. aince there weSmuch dog gnllwing on thMe parla. whieh deslroytl<1 anatomi
ca! p'opertias and eould have obscure<1b"tcherillll ml.ks.
cValuas were ilalidud if over 20% uf Ih.. elemento examinad ahowad butchllrlnll marks
'Elements w..r.. ollty Hate<1 when base of Intl ... was stillaU..r.hllrllo skulf
1 .14
2 05
1 .tn
O .00
O .00
1 .33
1 .50
1 .25
O .00
O .00
O .00
1 13
O .00
O .00
O .00
3 1.00
1 .17
2 .29
1 .20
o .00
4.5 1.00
J .33
4 .80
O .00
O .[JO
6 46
O .00
3 .50
1 33 2
6 1_00 4.5
5 .36 8
3 .43 5
O .00 3
1 .17 5
24 .73 13
fi.43 O
5 .42 6
o .00 3
4 .22 6
11 .42 7
6 .27 5
2 .10 7
9 .12 28
2 .to 10
::1 .13 12
3 .12 U
10 ,45 J
9 .19 2
1 04 4
2 fi5 13
2 .18 7
O .00 7
1 10 8
5 .25 8
O .00 36
O .00 33
O .00 32
%' MNF. MKD %' MNE MKD %.
(9) (10) uu (12) (13) [14/ (151
O ,00 3
O .00 6
1 .5U 14
O .(JO 7
O .00 3
o .00 6
1 .50 33
O 00 14
O 00 12
o 00 31
1 ..13 18
3 .75 26
2 .25 22
1 20 21
O .00 744
O .00 21
O 00 24
1 .06 2fi
5 .3J 22
::1 .4.1 31
O 00 25
O .00 43
O .00 11
O .Ofi 10
O .00 10
1 .50 20
O .00 toO
O .00 95
O .00 77
Rulland carbou Bear site cartbcu Bear sile sheep
L h-.-.......d the .fMmM-'m*rials <fmmthnw
rtW,d,enHaLenes. for" Iwo..speees, Dall sheep [Ovls
doJ/ij and caribou (Rongifer torandus). These are not
Cut Saw, and Chop Marks from Nunamiul
Eskimo Sites
00 1
.00 3
.00 4
.50 8
00 5
00 8
00 O
.00 1
.00 13
00 15
33 7
.00 7
.07 El
.[JO 3
.00 4
.00 3
.00 2
.00 6
.00 6
00 3
.00 o
1,00 O
:13 2
1.00 O
.00 O
.20 O
.00 2
.oo O
.50 1
(4) (5) (6) (71 (8)
o 00 O
O 00 1
6 ,46 1
::1 .23 2
2 .18 2
1 08 10
o 00 1
1 .25 1
1 .00 2
6 3(1 2
7 SU 3
1 .06 5
2 .10 15
4 ,40 6
O 00 4
1 .11 4
1 .13 3
O .00 8
O .00 8
O .00 6
O .00 O
4.5 1 (JO 3
2 .33 6
2 .67 1
O .00 1
2 .14 5
9 64 O
2 40 O
::1 .43 2
Kakinye cartbou
(1( Anatomical part
made in Alburquerque on ve assembleges of bone
collected in Alaska from behaviorally documentad
Nunamiul Esklmo sites. I will describe the lalter ob-
servatons Iirst , since I consider them lo he the most
behaviorally documented of the descrtpttve material
AntIer" 1
Skull 4.5
Mandible 6
Atlas 3
Axis 3
Cervical vertebras 14
Thoracic vertebran 14
Lumbar vertebras 5
Pelvis 7
Scapula 4
Proximal humeros 8
Distal humeros 13
Proximal redro- 13
Distal rado-cobttus 11
Carpals 12
Proximal metacarpal 5
Distal metecarpal 4
Proximal femur 18
Distal femur 20
Proximal libia 14
Distallihia 12
Tarsals 21
Astragalus 10
Calcaneus 10
Proximal metatarsal 9
Dis1al ml'!talarsal R
First phalange 42
Second phalange 35
Third phalanRe 31
muscle slripping, primarily wilh bone choppers, was
a basic dismemberment slrategy.
I have made pr;mary observations on two seis of
faunal data for {he purpose of isolaring and tabulat
ing cut marks or butchering traces on bone. The ear-
Iiesl observations were made in 1968 on !he fauna
from the Mouslerian levels at the rockshelter of
Combe Grenal excavatecl by F. Bordes and located in
Ihe Dordogne Valley of soulh central France. The
faunRI material flOm this site represenls the behavior
of Neanderlhal man between roughly 90,OUO and
-10,000 years ago. The second set of observations was
FIGURE 4.05, Cut marlu observed on whiteloil deer al
,he EscheJmon sife. fReproduced with perndssion /mm
Guildoy el al.. 1962,/
FIGURE 4.04. Cut morlcs observed on block beor ot 'he
Eschelman .lite. (Repi'Oducedwith permJuion from GujJ
doy et al., 1962.)
When man butchers 3D animal, he leaves traces of
bte cutttng and chopping aetivities. Few recera
studies hsve heen made of Ihese traces, although
such marks were B subect o early Dld World
In a study seemtnglv "befare i1slime," Henri MBr-
tin (1907-1910) considerad in detail traces o skin-
nlng. evteceeatton, dserttculaton. end llettng. as
well as the fracture and burnng o bones. He llus-
tratad nicely the traces made by stone 1001s 00 the
fauna from the famous Mouetenan site of La Quina.
These photographe remain sorne o the best pub.
lished exarnples cf cut marks made by stone tools.
{Sea, Ior Instanca, Martin 1907-1910: Figures XLIV,
XLVII, xivru. LJ.
In North America, the Irst systematic study of
butcherng marks thet 1heve been able lo loeate was
done by Guilday et 01. (1962) on Iaunal remains from
an Early Historie stte from Pennsyvente (the Es-
chelman site]. This study provdes sorne guidelines
Ior Investigating butchering marks: "To qualify as a
butchertng merk two cnterta were applied: fl} repet-
ition in spectmen after specmen al precisely the
same locatton on the bone: (2) thal there was sorne
analomically dictaled reason why a particular rnark
should occur at any given spot IGuilday et al.
Severet Quite useful conventtons o( presentation
and descrtptton were ntroduced in Guilday el al.
(1962). For tnstance. the use of skeletal silhoueltes as
a graphic wey of presenting data 00 bulchering
marks is notable. These aulhars were feirly syslema.
tic in presentiog quentitative dala from ther obser-
vatioos on lhe American black. bear. Fairly complete
quantitalive data were given (ar whitetailed deer and
pnn:tically none were glven (or elk. Figures 4.04 and
4.05 are reproduced rrom Ihis seminal sludy and l
lustrate the anatomical "hot spols" where butcher
ing marks were systematically observed.
Considerable allention has been given lo traces o
butchering by Georga Frisan (see Frison 1970). Most
of Frison's work has been with Ihe remains ofbison
al kili sites and hence should reflecl Ihe aclivities o
primary butcheting and processng for Iransport.
Fri!;on pioncered Ihe study of bone brellkage as ti
clue to butchering hehavior and has argued that
Butchering Marks

4. Human Mode-s af Horre Modificalion ButcheringMarks
heve listad vertebras or ribs but the absolute MNE
should not be treeted seriously only the percentega
modtcattons. Summary data on butcberlng marks
are gtven in Tabla 4.03. For additional comparativa
data see Parmalee (1965:25).
There is clearly pattern to the dim.ibu.
!L0n and frequencyof cut marks on bones at the.&.i.llL
Zf Combe Gr_enal. For Ihe Ihree specles examined
the ste contenta, since during 13 years of excavaticn
Bordes fol1owed traditional prccedures relativa to
the colleclion of fauna! materials, instructing his
crew lo save only teeth. articulator ends. and
in the beliefthat rtbs.Iong-bone spl intera.
vertebrae. srnal l fragmente of skult. scapula blades.
and pelvc parta other than the acelabulurn were not
sutable for spectes identification. In sorne cases, 1
TABLE 4.03
'1 ",' .. \ Dislribution 01Butcherins Murks on Bones Imm 'he Mousterion Site DICombe Grenul"

/ C",;'o;"
\l / primigenius EQuus Rongiferlatundus
1 :' __ Iaurochs] (horse) (relndeer
'L /'
Analomical (1] (2) {3} (4) (5) (6) (7) (6) {9)
par! MNE MKD %" MNE MKD %" MNE MKD %"
---- - ------_.__..-._---
_. .
f.:f' ."Jl1.1 Maxilla 2 O 4 _1
- MlIndible 3 3 100 7 5 86 Z7 Z3 85
,-: .. AlI.. 2 2 + 50
1; \ Axi5 Z 4
;)' ': I lnncminate 2 10 7 3 43
.::::-:. 4 61 6 10
. _., Scapula 1 6 2 33 44 4 9
',7 - ./,' Proximal humerus 4 4 1
,.' - -; 1. Distal humerus lO 1 10 7 1 14 25 3 12
Proximal radiocubitus 10 2 20 11 3 27 39 3 8
Distal radio-cubitus 3 2 25 1 4
Righl Carp..ls 3 Z 76 O O
FIGURE 4.06, Morks typically proolu:ed durinS the re- melacarpal 1 4 30 O O
moval 01lDeot lralD the scopula, Di51al metacarpal O 6 43 3 7
femur Z 17 3 18 27 1 6
This observation moves uso in the direction of dis- Distal femur 3 4 14
cussing the actual (orm and placement of cut marks. 2 2 9
a subject 1wish lo delay further discussion of unti! Dlslal lIbIa 7 9 1 11 59
. b d Tarsals 7 6 43
addilional comparatlve data have een presenle . A
slragalus 5 3 34 3 9
r..!cfmeus 2 40
Proximal metatarsal 2 6 1 13 48 1 2
Marks on Stone Tool Bufchered Animals from m61atarsal J 46 2 4
Combe Grenal Firsl phalange 12 93
Second phalange 6 90
The Nunamiut qs,emblages animals Thirdphalange 1 4 23
dismembered withmetaLlooJs.. It is quite reasonable
to Ihat butchering strategies woulcl be re- Total 68 6 9 147 17 12 931 63
sponsive to the charac!eristics of the lools available
d PI . ,. 'ded "MNE = moimal, numbar of alem..nts in Nchanatornical calagory not converted la MNls: MKD = numbar marked bycul marks.
for use. A goo elslocene companson IS provl "'alues wera ,Iahc'l'erl if o"er zn".. or tha",lem..nts!!xaminad shnwP.<l botchering marb.
by the fauna from Combe Grena!. The MNEs reported 'AlJ are Y<UlOg "spike" antlers wilh auachcrl s!!ctinos 01 skull. AH cut or "saw" marks around th!! ha.e of Ih., antier on Ihe
should not be taken as an unbiased representalion of skull, I hay!! observerl fmm Ih" mmoval af th!! skinnI Renerally lor USI' in c!othing manufacturE'
ell the essembleges 1 heve evaeble for sludy bul
they are all Ihal haya been studed in detall under
good lighls and laboratory conditions. Table 4.02
summanzes Ihe obeervettons on butchenng traces
from the five populations observad. Tha surnmary
treats-only the frequency-ef such marks, regardless
of-form.oo difierenl elemente of.the enetomv. The
infcrmatiun provdes a clue lo where lo lo'ok for
butchering merks, and aleo tndcatee the ateas of tbe
anatorny that were mcst cornmonly alterad or pre-
serve traces of butchering activity. Italicized per-
centeges highlight such anatomlcal ereas (values
were italicized if over 20% of Ihe elements examined
showed butchering marks].
Two factors are of nterest here. First, there are
between ihe butchering mark
distribution end the descriptions of butchering that 1
reponed in sorne detall. (L. R. Binford 1978b:48-6IJ.
In none of the field-butchering demonstrattons did 1
observe arry cutting or dsmemberment of the skull,
yet in these essembleges oH the skulls show sorne
evidence of havng been butchered tnto subunits.
Since the essembleges are all from residentiol sttes.
ths further processing of the skull must have taken
place during preperation for consumpton. Ths
same condtton is true for the pelvis. In the field-
butchering eplsodes 1observed. the pelvis was never
reduced to subelements, bul in all five assemblages
40-50% of pelves showed evidence ofbeing reduced
to subelements. The actual percentage is probably
larger, since most pelvic pans ended up in dog yards
on these sites and hence suffered cons.iderable gnaw-
ing damage. which may well obscure butchering
The secoRd item of interest is tha presence of lon-
gitudinal cut marks in the super- snd infraspinous
fosaae of aIl sheep scapulae from the 8ear site (Table
4.02 and Figure 4.06). It seemed c1ear to me that
these marks derived Crorofilleting the meat from the
scapJlla, an activity nonnally associated with drying
meal. (Sea the fillets hanging on the drying rack in L.
R. Binford [1978b:101, Figure 3.9j.) 1then examined
sorne of the scapulae from spring sites, such as
Tulugak Late Site 2A (Binford 1978b:2D5-213l,
where I knew drying was taking place, and found
that 83% of the scapulae remaining with blades ex-
hibited such marks, lending strong support to the
relationship suggested between this Iype of cul mark
and the dryinR of meaL
100 Humcn Modes oj " ...,T_.".-

... J

FIGURE 4.07. Morb by srone ools 00
tire tace of reinde"" mondlbJe. fFrum
the Mousterlan site uf Abr Van/re:;'.
excavoled by}. P. HigQud.}
FIGURE 4.08. Bu/cher romoving lit,. rOllsue {mm a cor;hou.
(80S primigenius, Equus cobnllus, end Rongifer
rorundus). the par! most cornmonly yielding cul
marks is the mandibie. Marks O" he IDllndible tend
to be slgbtly obltque incised merks on the inside of
the ruendible generally opposte the M2 tootb [see
Figure 4.07). The marks are believed to originale
from the undersde of the mandble and lo be related
lo the severtng of the mylohyoid muscle during the
removal of Ihe tengue. Figure 4,08 iIIustrates a
Nunamiut butcher going in underneath the mand-
bleand removing the tengue of a male certbou prior
to the removal of the mandible from the skull. The
next most common cut mark is around the lip of the
ecetabulum (Figure 4.22). Marks enctrclmg the
acetabulum are presumahly made unng the cutting
of the iliofemoral and tscbtofemoral ligaments,
which tend lo encase the coxal articulation. A minar
type of cut mark. observed is al rtght angles to the lip
of (he acetabulum and is believed to bave resulted
from the nserttcn of the cutting instrument into the
acetabuJum to cut the ligamenl of the fomoral head
afler the coxal articulation ts dslocated. Anotner
type of cut mark was conststentlv observed on the
distal humeros and on Ihe proximal radio-cubilus
.. ,

/ I

/ /
I "'{
/ I

I /
I /
I /
1 I .-_.....
1 1 , ,/ .1lH\
10 .!.c/T /.....c \
u 01OC1 ....., I
. .. .. .. .. .. ..
.. .. ...

FIGURE 4.09 Re'aUolIship betweell frequendes 01 cUI
marks 011 brmes Imm Ihe MOllsteriarf sitf! uf Combe Gre
nol ond lho'<c from thf! BerJr site 01 Anoktuvuk Po...c,
FIGURE 4.10. Murks of stone tO(lls011 base 01 reindeP-r
Qlltler. (Fmm Ihe MOllsterian site of Combe Grellal, ex-
covoled by F. Bordt'Ii.}
{Figure 4.31). This was a trensverse cut mark acrees
the lower part of the coronoid foesa and the olecra-
non fossa on the distal humeros. Cul marks were
common on the otecranon. generally nmning across
the procese. I will not describe all the cut marks ob-
servad al Combe Grenal here, stnce the point of pre-
senting the material is simply to demonstrete that
there ls enetomtcal c1ustering of merks.
Figure 4.09 compares the retndeer from Combe
Grenal wilh the caribou from the Bear stte as to the
relations between Irequency and plecement of
butchering marks. Blemente from Combe Grenal that
exhibil high frequencies of marks relative lo the Bear
stte material are the mandible, entler bases (Figure
4.10), and thescapula.l have already mentioned the
Mousteren mandibular cuts. As for the cuts around
the base of the antlers. it should be ponted out that
all the antlers observad in the Mousteriao material
were splke antlers ofvery young individuals. Circu-
lar cut marks were seen around Ihe base of the
antlers in over half of Ihe cases (See Figure 4.11). I
have previously noled that such cuts were mosl
common on Nunamiut siles occupied by spedal task
groups enRaged in fall hunling of young animals for
use as wlnler c1olhing. It should also be pointed out
Ihat the Nunamiul ski n the head with such eare only
4. Human Modns of Hone Modificolion Butcher-ing Morks
skin an animal by making an encircling cut around
the metapodiaJ or sornettmes the lower tibia and/or
radio-cubitus WIlS almosl always associeted wilh
bulchering aclivities al a mltiple kill , where the
primary concern was with meat , not skins. On the
other hand, when late surnmer-early {all hunts were
conducted specifically wilh an eye lo obtantng
yearling calves or calves o the year for ther skins, il
was 001 uncommon for Ihe butcher to take pains in
skinning out the Ioot. nittattng the nside leg cul
from the hoof. This would heve the effecl o render-
ing cut rnarks 00 the phelanges rether than the shafts
of lower leg bones. The butchers explained that the
lower leg skin served as essential componente in the
manufacture of mukluks (skin bootsj and skin socks.
The mainlenance of the natural conformation of the
skin al the animal's foot was essenel lo providing

FIGURE4.12. BUlcherins sr:hoof. mn Willer beins inslructed on how lu dsorcuiete fool from the melapodial.
when the skin is destinad for the manufacture of
where the .head of the cartbou is used for the
the parka
In a similar vetn tha very earIy study of cut rnarks
by Henri Marlin (1907-1910:229) reports considera.
ble numbers of marks described as "ineisions eir_
cu/arie des members." a cut mark encircllng abone
that was interpretad as having arisen in Ihe context
of skinning animals. This same tvpe of mark has
been consistently recogoized by modern workers.
The tnterestlng difference s that from the
Mousterian materials reported by Martin most cuts
uccurred on phalanges, whereas recent studies of
bson (Frleon 1971; wheat 1979) remens and other
forms show this to be common 00 metepodal bones
end relalively rare on phalanges.
In rny expertence with the Nunamiul, starung lo
Mork, from culling off the heod, 5-1
( b)
Morks ftom skinnln;. 5-4
nGURE . 11. Cul marks observed 011 skulls: (o) PQslerior, ond (bJ ventral
, '
4. Humen Mcrtes of Bane Modificotion CutMarks: Their Formnnd Placement on SpecificBones 105
1: I
the mplement used lo accomplish the tesk. In both
the form end placement ofmarks I hope to document
vertabtlty thet may be reliably referred to known
forms of butchenng ectfvtty as wel1 as to known
forms of tools. Of course what is desired is the speci-
ficalion of dagnosuc characterstcs of cut marks
that may serve as the bests for robusl tnferences
ebout the charac!er of past acttvtttes and types oC
The faunal materials I colleoted among the
Nunamiut Eskimo were from animals butchered
with metal tcols. The data I collected from the
Moustertan site of Combe Grenal are reCerable to
bulchering actlvites conducted with stone tools.
Unfortunately, al rhe time 1collected the Mouetenen
data I was not very knowledgeable about butchertng
and waa not aware of the propertes of cut marks
made with metal tools. Thls meens that a well-
informed and comparebly collected body of data that
mighl elucidate derences between stone end metal
tools s not available at presento
I heve made cerfain observetons that mey etd in
differentialing the two but certainly more reliable
entena of recognttton could be develcped if control
materials were available (see Walkerand Long 1977).
Masl of the cut marks made on bones with melal
tools are almost,hiirl1i:ie in size. They afien appear to
haYa been cut irttt'llhe bone from Ihe side, or oh-
liquely.leaving an overlapping small "shelf'" oCbone
tha! romains in place. For this reason cut marks pro-
duced by metal knives are very hard lo .'lee when one
looks direclly down on abone (see Figure 4.14). Of
course this visibility varies with Ihe pressure applied
and is probably an inverse function of how well
honed Ihe knife rnighl be. In addition lo having Ihis
diminutive "sliced" look, Ihe marks are generally
long, resullng from cuts running across Iissue for
considerable dislances. Cutting with slone tools re-
quires a much less conlinuous aclion, more of a se-
ries of short paraUe] Also. mosl stone tools,
particularly ones that are relouched. do nol have
straighl or single-plane cutting edges. Marks from
stone tools !end to be short. occurring in groups of
parallel marks. und lo have a more open cross
seclon. Thev also have a more ragged appearance
when viewed from Ih!" topo Cut marks undoubledly
Cut Marks: Their Form and Placemenl on
Specific Bones
The morphological selting and !he form of a cut
mark will vary sornewhat wilh Ihe type of work
belnR conduded when Ihe marlo: was produced and
posterior tace into the lotnt between the articulator
surtaces of the two bones end severng most of the
Igaments connecting the two bones (oblrque end er-
cuate pcpltteel. anterior and posterior cructate. an-
terror and posterior menscofemorel llgaments). This
strategy leeves a pair of telltale Iraces: a sel uf
semicircular cut merks around the lateral tntercon-
dylar tubercles of the proximal tibia {cut Tp-t . Fig-
ure 4.Z6) and cut marks on the medial surface of the
lateral condyle of the distal femur. The Iatter marks
sometimes exlend partway inlo the ntercondylar
fossa {cut Fd-3, Figure 4.25). In a similar way. the
dsertculeton of the oot from the distal metepcdtel
is eccompliehed by the Nunamiut by lnseeting the
polnt of a melal the nexed joint. as shown
in Figure 4.1TNicis$ on the posterior condylar
spines result. Thiftechnique of inserting a knife
point into the [olnt is probably inappropriate for
stone tools snce stresses are set up on the flat sides
of the blede. Such stressas would render a snep break
likely with storte tools but not with metal
It is my impression thet with stone tonls Ihe most
common stra!egy couples more superficial cuts wtth
fhe application of leverage lo dislocate a [oint. as ls
tllustreted in Figure 4.13. By ustng a dtslocancn
techntque. less scerrtng of the articulalor ends of
bones is Iikely. The Hgaments are stretched end ex-
posed. so they may be cut directly. Perhaps this con-
trast accounts for sorne of Ihe othar inflaled cut mark
frequencies on articulalor surfaces of long bones in
the Nunamiut malerials.
Thus Car, I have been speaking primarily of the
distribution of butchering marks on Ihe skeleton:
however. Ihe foregoing discussion leads naturally
into a discussion of the formal properties of the cul
marksand their specific placemenl on bones. Befare
taking up this issue.1 should emphasize Ihat a popu-
1alion of moderale-sized animals butchered by man
using stone tools can be expected to yield evidence
ofsuch butchering in the form of cut marks, regard-
less of Ihe Iypes of culting lools used.
, :
..... ..' \ .

lary dental ares, rendering examinatlon Cor this cut

mark impossible.
Th8-Moosferian materlitl'
des, of .cut.marks-areumithe' head af,tM HElJRlla.
Such IDa,," were-l'a1'8 orab&anl {rom the Nunamiut
material, but eoromon on bear bones frQnLthe Es-
chelman site reported by GuHd8Y el al. (1962: Figure
2). The meaning of Ihis difference is not e1esr lo me.
Bones in the Nunamiut material that exhibil higher
frequencies relative to Combe Grenal materials are
dislal femur, proximal tibia. proximal and distal
humerus, proximal radio-cuhitus. and distal meta-
tarsal (See Figures 4.25, 4.26, and 4.32.) Al least
sorne of these difference.'l are in my opinion relatad
to the use-of melal tools in hutchering. For nslance,
Ihe procedure for disarticulating the femur from !he
tibia involvas inserting Ihe poin! nf a knife from !he
FIGURE 4.13. Butcherlnll 5chool. Don WJnel' bejn8 Instrucled jn the use olleverag!! as o butchering oid.
the propar shapes fOf use in farmiog lbe instep area
of the sock Of boot. Pechaps there is one 8uch similar
concern standing behind the time-consuming ac
tions of skinning out tha fool indicated by (he
Mousterian cemaios.
Other characteristically noted on the
Nunamiut bones relatad lo Ihe carnoval of Ihe head
&oro Ihe neck. TypicaJly lhere were marks across
the ventral sunace of the occipital condyle (see
Figure -4.11) as well as related transversa. cut marks
aerass Iha ventral surfaee of the anterior wings
of the atlas vertebra. 80th marks derive from sevar
ing Iha head from the neck fsee L. R. Binford
t978b:354, FiBure 7.11}and are very common on Ihe
Nunamiut tnalerials. The frequency among
Mousterian malerals s unknown since the only
parts regularly saved from the skull wen' Ihe maxil
t + rt.; ,'.:1- '1-
t\ ("" .}Ii:> ;,,;
, -, ...
(1J-;) :; \:. t;':' 1.
.. .
I! h..;
." l,"" '"
. .,.., ,..t- ..
H'r ar .,.,. J
1 (J.1l_" ).,""

FIGURE 4.14. Chal'Octeristic marks produced by metal
knives Ilsed In blltcherin,l,
produced by stone tools are shown in Figure 4.15,
whera abone from the Mousterian site of Abri Vauf-
rey (excavated by J. P. Rigaudl is shown.
Placement of Cut Mllrks
Many have recognized that cut marks derive from
different stages of processing en animal. Since such
slages normal1y resul' in an accumulation of marks.
sorne traces of earlier snd more fundamenlal slages
of processing remain on Ihe bones as they are sub
jecled 10 more exlensiva processing during Ihe se-
quence of acHans beginning wilh the procuremenl of
theanimal and terminating wilh 'he abandonment of
Ihe lasl elemenls of the skelelon.
This sequence is almosl ahl.'a}'s{al skinning, (bl
dismemberment.l el filleling for either consumplion
4. Human Modes of aone Modification
or storage, which normally involved sttll further
dismembermenl, and {d] marrow consumption. It ls
true that marrow consumption may accompany
skinning and initial dismemberment, butthis is rec-
ognizable BS consumption of parts Ihat yield marrow
but little else, such as lower Iimb bones. Marrow
removal fram meat-yielding bone norrnally takes
place aer the rernoval of meat ter consurnption or
storage. This normal sequence implies that cut
marks eccumuleted 00 bones will vary with the stage
of processng reeched prior lo abandonmeot of the
bones and will vary in dicgnostlc ways regerdtng the
plecement of en assemblage in a logistical-
consumption sequencB.
Since I am eoncemed with documenting prop-
erttes thet may be investigated foc their diagnostc
valu in seeking a behavioral knowledge of the pasto
it ls perhaps useful to describe in generic terms the
types of cut marh Ihat can be expeeted from dif-
ferent stages of processing of an animal by mano
Skinning Marks
There are actuaUy very few places on Iha anatomy
where Ihe manipulalion 01 the skin brings Ihe
FICIlIRF. 4.15. pr(lfluc'd by slane
loob in bUlcherin".
Cut Marks: Their Form and Plncement 00 Specijic Bones
butcher in direct conteo with bone. The two places
where this te mosl Iikely are the lower legs and the
lwu!.. I have already mentioned that cuts derived
from skinning actvtttes have been noted enclrcllng
the shes of lower Itmb bones. Such cuts have been
observed on the lower tibia, the shaft of the metatar-
sal (Figure 4.16J, and the phalanges. Analogous
placement has been noted 00 the front leg, the distal
segment of Ihe shaft of the redo-cubtus. tha
metacarpel, and the phalanges. This encircling cut is
generally one of the rst to be made and is basically
the "point of enlry" for the skinrung of Ihe animal.
Subsequent cuts normally originate al such en en-
circling cut and extend down the medial face of the
leg to the body of the animal. Cut marks may result
from thts action.
The second place where the skin is essentally
stretched over boDe s on the head. There are severa]
places where one might expect cut marks if the head
is in fact skinned oul: around Ihe base of the antlers
(Figure 4.1 t) or horns, somewhat less commonly
around Ihe eaes, and around the moulh. particularly
in Ihe "chin" area of Ihe mandibJe.
As noted previously. skinning for skins differs
&om skinning as a stage of butchering. WheD Ihe


Il /MTd-'

FIGURE 4.16. Marks indirall-r. af likinning activities.
skins are destinad for use in clcthtng manufacture.
the animal s skinned out fairly ccmpletely. resulting
in cuts around the phalanges. around the entlers.
and on the chin area of the mandble (sea frisan
1970:11J. These basic marks are ndtceted as No. 1 in
Figure 4.04 and Nos. 24, 1, and 3 in Figure 4.05.
Good ilIustrations of Ihe cut marks in the phalangeal
ereas are given by H. Martn {1907-1910}: Platea
XLV. No. 1, Plate LIX. Nos. 1. 2, 3,4, and Plate LlX,
No. 6 show an enctrclfng cut eround a metepcdial.
Dismemberment Mcrks
By fer sorne of the more distindive cut marks de-
rive from the dsmemberment phase of processing an
animal by tool-using mano In most cases dsmem-
berment consisls of disBrticulalion; hence, cut mBrks
are associated wilh points of articulation.
I have observed ooly two Iypes of cut marks that
may be exclusively referred lo the dismembennent
stage of processing. The most common, as well as the
most reliably referable to the dismemberment stage,
are marks remaining from the removal of Ihe heBd
from the nock aOO from lhe removal of Ihe mandible
andior tongue from the skull. Figure 4.11 iJlustrates
Ihe characteristic transverse cut marks on Ihe ventral
surface of the occipital condyles (S-l). Correspond-
ing transverse cuts occur on Ihe anterior ventral sUr-
face of the atlas vertebra (Figure 4.20).
In sorne cases the head seems to have been re-
moved by Bcting on the articulation between Ihe
atlas and axis vertebrae. wilh this stralegy marks
may occur on Ihe posterior ventral face of Ihe atlas
and on the anterior ventral face of the Bxis.This pat-
lem is well ilIustrated by Frison for antelope
(197t:264 Figure 3aa-dd) Bnd is indicated by Henri
MBrtin (1907-1910: Plale XLVII. No. 61 for the
Mousteriao site of La Quina. In my experience this
melhod was employed by Ihe Nunamiul on frozen
carcasses or 00 aoimals Ihat had remained in Ihe
field focsorne time between Ihe tima of death and the
inHial bulchering and had therefore become very
stiff. lInder Ihe lalter conditions lhe most mobile
joinl is Ihat belween Ihe alias and axis, snd "break
&A :5) "(f "'h
" I
, i
.cJ.; \
FIGlJRF. 4.18. Coribou dw/l cuf in hal! ond wilh nQSC removed.
FIGURE 4.17. Caribou duY sbowlng the
mus' common method o/ removing the
4. Human Mcdes of gene Modificaton
Cut Morks: Their Form cnri Plocement un Rones
tng the head loase" by pulling it around using the
entlers as a lever is sure lo be easer when the axis-
atlas [ont ts penetrated.
On occaston. primary butchenng may ndude the
removal of the antlers. particularly if tha head is lo
bereturned to camp and limited rneans of transporta-
tion are avetlabte. This task can be rather formidable
if ene does not have axes. saws, DI other heavy-duty
tools. In the abeence of such 100110 Ihe cranium ls
usually broken around the base of the antlers and the
antlers are removed. Breaking the skull case is rnuch
eester than atlempting lo cut through the antlers.
Figure 4.17 illustrates the resulte of ths stretegy. AH
the antier bases reccvered from the Mousterlan stte
of Combe Grenal exhiblted edherng segmente of the
ekull, ndcatng that thay had been removed this
way. The Nunamiut frequently place heeds in caches
and segment them into perta appropriate to cooktng
on removal from Ihe caches. For fall-killed canbou.
(he head is nurmally frozen when t is butcherad. so
the Eskimo use s SItW lo split it nto two parts. The
nose of alarge bull mighl be cut offhefare Ihe head is
split snd prepared independentiy and differenlly
froroIhe split halves oflhe head. Figure 4.18 shows a
bull hesd sawed in half righl down Ihe medial tine
from front lo back.
Since many heads are secondarily bulchered
when frozen, thera is a distinclive mark left on Ihe
skull from a particular cut made during Ihe removal
of lhe mandible when Ihe entire unit is frozen. Since
Ihe head with attachad mandible is complelely stiff,
a deep and long cul is made from Ihe insert of Ihe
masseler musde along Ihe upper lip area aboye Ihe
upper molars direclly baek aeross the ascending
ramus ofthe mandible. severing the masseler muscle
oompletely. Once Ihis is done the mandible may be
manipulaled slightly and Ihe lask of removal is
made much easier. Figure 4.19 shows a bull caribou
skull with cut-off nose, and Ihe distinctive deep cut
aboYe and parallel lo tha upper molars.
There are two seis of cul marks Ihat may Dccur
Somewhal independenlly of one enolher. Slrictly
speaking, Ihe disarlil:ulalion of the mandible from
lhe skull is Ihe anly dismembermenl ae!: however,
temoval of Ihe tangue is commanly execuled during
initial butchering retber than during meat dislribu-
tion or al the stage of preperatlon of parts Ior con-
sumplion. The Ierger the animal, or the more rtgtd
the carcass has become. the more difficult t s lo
remove the mandble. Therefore, more cut rnarks are
generally assoceted wilh larger animals or with
animals of almosl eny stze that heve not been butch-
ered immedialely after death.
For nstence. tbe moose jaw shown in Figure 4.14
is from the same animal shown during butcherng in
Figure 4.44.1 observed in great detall the removal of
tha mandible from ths animal as compared lo analo-
gous removals from caribou. Whal dtstrngulshed the
removel of the mcose mandble was the process of
fiJIeling in sttu. The masseter muscle was cleened off
the mendfble as well as stripped off tts insertion
ponts on the skull. This resultad in cut marks
eround the edges of the massetertc fossa and lon-
gitudinal cul marks completeJy across the fossa. In
conjunction with these cuts was a diagonal cut on
the maxiJla just behind Ihe third molar (sea Figure
4.19 and Frisan 1971:264, Figure 3ft). Once the
majar muscle had been removed, cuts were made
againsl the inferior surface of Ihe mandibular con-
dyle, Than the mandible was mechanically manipu-
Jaled and essantially pried loase from Iha skufl with
small cuts of connective tissue accompanying Iha
pulling away of the mandible. This seems to be Iha
same procedure described by Frisan (1971:265, Fig-
ure 3, w-z, ff) fm butchered antelope from lhe
Eden-Farson site, Wyoming: "The axtent and depth
DE Ihe culs do not appear to have been necessary to
cut Ihrough Ihe jaw musclas alane [Frisan
1971:2631." The saemingly "unnecessarily" deep
culs, the concenlralion of Ihe cut marks al the axjs-
alias erticulalion of Ihe neck, and the large-animal
strategy of removing the mandble leed me to sus-
pect Ihal primary butchering look place afler rigor
mortis was advanced ami/or Ihe animals were partly
When the longue is removed while tha mandibla
iR slm aHachad lo tha skull marks are usually in-
nieted on Ihe medial margins of the mandible jusI
below Ihe Ihird and fourth premolars. Thesa are gen-
erally produced in conjunclion with severing Ihe
mylohyoid muscle, which tands to free Ihe lateral
margins of the longue so it may be cut free al ils
"rool" (see f"igure 4.08).
, ~ " "
FIGURE 4.19, Corlbou skuJl wlllt aese removed ond distlndJve burc1lerlns mork obove Ilte molors.
Cut Morks: Their Form cnd P/ocement on Speci/Jt; Bones
metal knves with long (6-inch) blades. This optnton
is besed on an nccrrect idea of the butcherng pro-
cese. since the roas! is not cut loase from the nb
atlachments but Is instead stripped along the sheath
of sinew as 1 have illustrated (L. R. Binford 1978b:
Figure 2. t l. It is further besed 00 en Incorrect idea of
the tools usad by good field butchers. The Eskimo
use the terms young men's kmves and old men's
knives. Young man's knives are considerad nonfune-
tonel for butchering and rether silly, being selected
by young boys returnng from boerdlng school who
have heen impressad by Euro-American culture in
the "lower 48" but know nothing about butchering.
These are the long-bleded kntves Spiess imagines in
use by the Nunamiut!
Given Spess's opinion it s rather ironie that the
cut marks origtnaung during the course of removing
the tenderloin are sorne of the most common merks
recorded. I have observed them on thoradc vertebrae
from Levels K, L, and 52 at Combe Grenal; Henri
Marhn /1907-1910) reports at leasl ooe spedmen
from La Quina. They are cited in almost all tha re-
ported assemblages where cnt marks were noted
from prehisloric bison hunters' sites on the American
Plains [sea Frison 1970:22. Figure 14; Frison et al.
FIGURE t.zo, Cu' marks an 'he
atlas vel1ebro produ&eri when sev-
ering the head from .he IIeck.
Clll marks ev'Z
Clltmarkt eV-1
Ventral yje'N
slabs from the spnal colurnn (TV-5j. Witb rnoderete-
sized animals the segmenteon of Ihe venebree
normally takes place between the second and third
thorectc vertebrae and the thirteentb and fourteenth
Ihoradc vertebras (cut 18, Figure 4.05). In most
cases this segmentation tekes place eer tbe re-
moval o tbe tenderlotn or eye roest. (See L. R. Bin-
ford [1979:49 Figure 2.11 for en illustretion of the
tenderlotn betng removed.) The strippng back of Ihe
tenderlotn is made possible by long cuts made along
the body of Ihe muscle, freeing it from the dorsal
spnes (Figure 4.21, TV-2) o the thoracc and lumbar
vertebree. as wellas from the superior surfeces of the
,ios (Figure 4.05, cut 20) end ther erttculettons with
the vertebrae or the transverse processes of Ihe lum-
barvertebree. Cut marks resulting from the laller op-
eration are commonly oriented treneversely or
slightly obliquely lo the dorsal splnes of the thoracc
vertebree. These transversa cut marks are shown in
Figure 4.21, as are a minor set of modifications that
resull fram inserfing Ihe knife and innichng marks
00 the dorsal spines between the verlebrae during
Ihe segmenlation of the spinal columo.
Arlhur Spiess (1979:291) has suggestcd lha this
method or bulchering is made possible by Ihe use of
4. Humen Mcdes 01 ucne Modi/icolion
There are basically three operations that may re-
sult in marks on Ihorade vertebrae (Figure 4.21) dur-
ing primary butchering: segmenting Ihe vertebrae,
filleliog Ihe tenderloio (TV-2). and removing the rib
tebra andlor lite sixth cervical vertebre. Most such
marks are inflcted during secondery butchering ec-
companylng meat distribulion or during food prepe-
Aside from the cut marks ncted during tbe re-
mcval of tbe head from the neck (Figure 4.20), little
dismemherment of tbe neck proper is conducted
duriog the primary butchering of mosl animals.
Sorne marks may be inflicted on tbe sixth cervical
vertebra if ibe neck is butchered off from the spinal
column {see Henri Martin 1907-1910: Plate XLVI,
Nos. 6 and 7]. When frozen carcasses are being
butchered, an ax or othar massive tool may be usad
resulting in chop marks somatimes al Iha axis ver-
> -





e on

U: t- -s




,. .
Cut Morks: Their Form and Plccement on Specific Bones
1976:53: Otlbert 1969:290; Wheat 1979:67, Figure
As prevously lndcated. marks somettrnes occur
along the trensverse processes oc at the base of the
dorsal sptnes. These marks derive from Ihe removal
of the tendertotn (see Wheat 1979:67, Figure 33h).
There are generally three locetlons for cut rnarks
inflicted on Ihe ribs and stemum during primary
butchertng. Transversa marks, derived from the ce-
moval of the tenderlotn. oceur along Ihe dorsal SUf-
faee of the rib just [o the side of the proximal end of
the rib (cut 20 in Figure 4.05). The secand most
eommon mark resulte from cuttlng off the distal end
of a rfb during tha dismemberment of the stemum
frcm the rbs. The last place where cut marks may
occur ts across the ventral surface of the rib, clase 10
the proximal rib head or sometmes across the ar-
ticulator bead. This cut derives from the removal of a
rib sleb from the spinal column. Normelly the
rnethod te lo enter the rib cage by severing the distal
ends of the ribs from the brtsket. Attachmenls are
relatively soft and cartilaginous in cherecter and may
be reedtly cut with even a dull instrument. A cut is
then made down between the seoond end thrd rfbs.
Fcr a detallad descriplion of this butchenng step see
L. R. Binford (1978b:53, 54, 95). Next the slab of at-
tachad ribs s pulled up sberply. generally resultng
in breaking off the rib haads egetnst the vertebree.
Once the rtbs are snapped back. the knife is ron
along the break severing ttssue along the ventral sur-
face. It s this acuon that resulte in cberectersttc
transversa or slightly oblique mark.s un the ventral
surfaee of ribs very clase lo the proximal erttculator
end. Ths ls an andenl technique: Henr Martn
[1907 -1910:233, Plate LVm, Nos. 10, 11, 12) de-
scrtbed in detetlthts exact cut en many rtbs recov-
ered from the Mousterian site of La Quina. I observad
ths cut on 10% 01' the nb heeds saved at Combe
Crenal. Il IStnteresttng that this mark is nol reported
frOlTl lhe Norlh American Plains siles.
In rny expenence. most of the marks that regu-
larly occur on the pelvis or the saerum [Figure 4.22)
derive from secondary hutchering C8.ITied out during
"."\ ..Jtl' ./

Venlral vi'w,

FIGURE 4.ZZ. Cut marJcs choMclerislk.eJlly prodlJced an
the pelvis during dismemberment.
.$r"'''1lr ..
.3 1:
'a\ i

il i
114 4. Human ModesofBone Modificotion
Cut Marks: Their Pormond Placement en Specifk Bones
FIGURE 4.23. Eskimo bukher removins fhe rear IflR010 curibou rull in 'he monDer fha' lem'es ,hr. m a r k . ~ .'ihawn in
Fi,lure 4.24.
FIGURE 4.24. Cur m a r l ( . ~ on Ihe pelvis produced while di!imemberinS Ihe rear le, /rom the pelvis.
muscles from Ihe bone: the insertion is used as e
handle to pull the body of the muscle from the ani-
mal. Little cutttng wth a knife is megtned when this
proposed procedure is tollowed. The argumenl hes
been very widely accepted (see Agenbroed 1978;
johnson 1977, 1978: Wheal 19791 as a documented
"fact" of North American Plains Indian behavior.
The procedure as descnbed by Frlson Involves the
slripping of rouscles from the rear leg and the ehop-
ping off al inserlons on Ihe pelvis prior lo the disar-
Iiculation al Ihe femur-pelvis articulation. If that
was Ihe case. Ihem would be no need to make Ihe
when tbe secrum ts dtsartculated from the pelvis
during primary butchertng (most eommonly done
with lerge anirnals], there may be a Irirnming of the
lateral margins of Ihe sacrum. resulung in cut marks
longnudtnally clown the iliae wings (PS-1 Table
There has recently appeared in Ihe Americanisl
llterature considerable dtscusson o butchering pro-
cedums. Ihe mos! influp.ntiai of which is certainly
the work ofCeorl'le Frison (1970, 1971, 197E1]. Frison
has arKued lar ti musde-slripping procedure, in
which choppers are used lo fr>c the inseMions of
ing bone, while orientad Iransversely to the ventral
surface of the pelvis jusI anterior lo the aeelabulum
[the move resulta in cut PS-7, Figure 4.22). An anal-
ogous cut commonlv begns in the rear along Ihe
"potnt" produced by the extenston of the isehial
tuberosity and runs obliquely down and anteriorally
over the "arrn'' of the tschum. lmpaetng the bone
transversely on the Ischel body just below the mar-
sin of the aeetabulum (cut PS-8) (See Figure 4.23.)
These Iwo initial cuts (Figure 4.24) may then be
supplemented by tlssue-severlng euts eround the
drsloceted femcral-pelvc arttculetton. resulting in
enclrcllng marks around Ihe ecetebulum andJor en
the heed of the femur, or more eornmonly on the
greater trochanter o the proximal femur. Sometimes
the cuts may be made inside the ecetebulum in arder
to sever the femoral ligament.
meet distribution oe during the processtng of perts
far etther storage ur consumption. Marks Inflicted
duriog primary butchering are generally concen-
trated just anterior or posterior of the acetabulum
socket on the ventral foce o the pelvis. These marks
are prcduced duriog Ihe removel o the rear leg from
the axial skeleton. The animal s commonly resung
on its back ur slighlly turnad lo one side. The foot o
the butcher te pleced in the crotch o Ihe animal end
the leg s then pulled and lwisted over eganst the
foot. resulting in the dslocaton o the artculation
belween the head o the femur and Iba ball socket
o the pelvis (acetabulum). Once the leg is dislocated.
the knife is usad lo cut oblfquely into the flesh from
the obvous fossa between the tissue o the abdomi-
nal wall and the muscles of the leg. The knife Is run
nbliquely down and lo the reer. frequently contect-

118 4. Human Modas of Bone Modification

Cut Morks:Their Form ond PJocemenl on Spedfic Bones
FIGURE 4.l5. Morks pl'Oduced on Ihe ff'mur durins dismemberment.
(. )
Lolerlll side

Anterior vie.
Ois/1I1 riQht
Yenlrol view
P051erior Yie...
pOSlerior Yie.
0;'101 riQM,
Circular parallel marks around he intercondvlu
tubercles (Figure 4.26dl of the proximal tibia-are
produced al lhe same time as the marks on the inler
nal face of lhe distal femur (Fd-3), Thal is, a knife is
inserted from the rear of the Hbial-femoral joiot and
rotaled around the inlercondylar lubereles (Tp-1),
simultaneously producing marks around the tuber-
eles and on Ihe interrlal face of Ihe lateral epicondyle
of the fernur (Fd-3). This mark (Tp-Il is shown in
Figure 4.26, b snd d.1t is mosl often produced when
butehering is done wilh mpta! knives, bul at leas!
one example has been reported from a prehislmic
North American site, the Jurgens site (Wheal
1979:56). This QCCUrTence can be taken as clear el'i
dence that disarticulation can be accomplished wilh
slone lools by inserfing the cutting implement be-
(wcen the articulalor surfaces uf the tibia femur.
The only olher cut mark re;:ularly noled on Ihe
tions. Cenerelly. tbe proximal libia and the distal
fmur rometn articulsted after inlttal fip.Jrl butcber-
ing end prior lo meat distribution or the preperetton
of meet ter eonsumption. Under sorne spectc condi-
uons. where there are euher large body sizes or
wbore rnultipie idUs incrl?Bse the bulk of eld-
butcbered meet that musl be trsnsported. dsarticula-
tion of tbe femoral-ttbel erttculanon may be accom-
plished during early butcherng stages.
Tbere are bascslty two general loceucns where
cut marks occur on the distal end of the femur. There
may be a mark oriented aeross the posterior surface
just above the lateral and medial condyles (Figure
4.25dl, A seeond and somewhat more ccmrnon loca-
tion IS across the trochlea or the peteller surface OJl
the anterior face (Figure 4.25e), These merks arise
from the nsertton of a knife behind the patelle with a
cut generAlly dtrected downward. so tha( (he patella
is removed with the proximal tibia nsteed of the
distal Iemur as ene rntght imagine.
Releted to the dsarticulaon of thts [ont with
metal knives as descnbed earlier, a characteristic
mark [Fd-3) is produced on Ihe medial face Df the
lateral ecmdyle (Figure 4.25f). This is general1y pro-
duced when the poi ni of a metal knife is inserted
inro th.e join! and !urned around the intercondylar
lubercles of the proximal libia.
lo many cases the removal of the rear quarter
couplecl wilh the removal of the lower limb is Ihe
only Belion teken during primary hulchering. In al!
my experieoces with he butchering of animals, even
relatively large aoimals, the rear leg Is dislocaled
from Ihe hall sockel al Ihe acetabulum prior to the
insertion of the knifo and Iha severing of Ihe
tive tissue between lhe mea! of Ihe t1pper leg and !he
musc1e atlachmenls in the pelvic area. It is during
Ihe lalter operation Ihat marks may be ioflicled on Ihe
neck of the femur (Figure 4.25a, b, cut Fp.l), 00 thf!
hall of the femoral (Fp-2), encitcling Ihe margin
of lhe femoral head (Fp-3). as weJl as on bolh
lrochanlers. Marks on Ihe grealer Irochanlerare more
eommon (Fp-5).
Cul marks 00 the dislal eod oE the femur are most
often prmluced dUling secondary butchfJring ac-
Wilh only two excepttone. the h811d of the Iemur W!lS
nlDl.Oved Irom the acetebulum. This is no! loo rliffkult
wth the flash stripped awey. The leg need only be glven
1:1 sharp jerk usjng the hump formed by the chnpped-eff
trochanter rnajor as a Iulcrum lo lever the head of the
femur from the ecetabulum socket IFrison 1974:411.
In all faimess, the merke dtscussed prevcusly are
not dentted al the Casper ste. Nevenheless. Fri-
son'e modal Foreccounttng for the dtseruculetons of
the femur ts what he had in mind ut the sitos where
he marks are present. Needless lo say, 1coMider
Ihese marks (;onc1usive eviclence that muscle strip-
ping as describecl by Frison was nol being con-
dueteJ. I wll treal in more de/aH lho evidence for
musc1e slripping io cliseussions lo follow regarcliog
choppiog techniques and breakage patteros.
cuts that, as 1 have indicated, would yield marks
PS7 and PS8 as well as PS-9, slnce all these marks
derive from severing the muscles 01 tbe rear leg dur-
ng the act of dsertculattng the head of tbe femur
from the ecetebulum. In Ihis regerd it ts tnteresttng
tha! tbese mares are dccumented al a number of the
sttee where muscle stripping is sad lo have been the
dominant method of butchertng (Frison 1970:16,
Figure 8a-b; Wheat 1979:66, Figure 32, lateral view].
Frisen comments on the leck o ertculetton betwecn
the pelvis and the femur al the Casper site as follows:
,,{.I '\',
", <,' ,'o VMe,,! I ,u I ,,,el
'\:'.X'l\? (1 UeL;
",,';;>\('" ')1, \
fr' ~
As has beeo indicated, encircling marks atong the
margins of the proximal end o the metatarsal can be
et the tuber caleta or the posterior end of the cal-
caneus. In my experences with the Navajo, where
the carcass te hung [see L. R. Binford and [, B. Bar-
tram 1977:92-93), and with the Nunamiut where the
completad upper rear leg is commonly hung on dry-
ng racks in cool seasons. enother action is belteved
to produce cut marks on the dorsal ridge of the cal-
caneus. In both ethnographc cases the tissue be-
tween the sha of the tibia and the ettachment of lhe
tendon at the tuber calcis ts cut with a knife to fadj-
tate inserting a rope or a gambrel for hanging the rear
leg r the cercees for further butcherng. This cul
"for hengtng" was observad to resull In marks illus-
trated here as TC-3 (Figure 4.27. b and el. lo light of
ths cbservetton it is interesting that this mark is
regularly reported 00 bison bones Irom the Jurgens
ste (Wheat 1979:66, Figure 32, medial view), and
antelope bones from the Eden-Farson site (Frisan
1971:284, Figure Jr).
An alternative method of disartieulating the tibia
from the metatarsal is to make the cuts at the nter-
Iece between the tarsels end the proximal metatersel.
whtch results in marks trensversely orented around
the joinl, scarring the medial surfece ecross the
intemal cunieform or on the medial margin ol the
proximal metatarsal. On the anlerior face, marks may
occur on the navicular.cuboid (Figure 4.28, TNC-1)
as well as the anterior face of Ihe intemal cunieforrn
lTE-l) and the m a r ~ i n of the proximal metatarsal
lMTp-1). Laterally, marks may occur on the proximal
articutator margin of the proximal metatarsal and Ihe
lateral face of Ihe navicular-cuboid. These marks
may frequeotly occur together with Ihose just de-
scribed, since the removal of Ihe metatarsal for mar-
row processing commonly involves the rernoval of
Ihe aUached tarsals, as is shown in Figures 4.02 Bnd
4.03. The combinatian of marks from both points of
disarticulation is likely to be a good clue lo al least
some of Ihe functioos ofthe site, since the processing
of marrow bones is normally associaled with served
meals rather than sna"king in hunting camps and the
Iike as is the case among Nunamiut. Similar dif-
ferences in processing may well characterize other
peoptes as well.
CutMarks: Their Formand Placement on Specificsones
The foregoing description touches upon al! Ihe
common marks observed on larsals (Figure 4.27) ex-
cept one, cul TC-3, which appears on the dorsal sur-
face of the calcaneus just posterior to Ihe point (lf
articulation wilh the astragalus. This mark has been
frequently intArpreled as deriving from the cuUing of
the lendon attachmenl of lhe gastrocnf!mius muscle
proximal tibia runa transversely across the posterior
margin of both the lateral end medial condyles [Tp-
2). Thts merk is observed relavely rarely and it may
be conedered as an alternativa to the cut thal runs
across the condyles or just aboye them 00 Ihe post-
erior face of the dislal femur {Fd-f , Figure 4.25d}.
The distal tibia appears te yield a very redundant
pattern of tnfltcted marks regardlees of time period
or geographical location. The marks I will describe
bave been reported from every faunal assemblage I
have consulted, from the Mousterten to the Historie
perlad and from Europe as wel1 as the New World.
Byfar, Iha most common mark s illustraled in Fig-
ure 4.26 (e and f, cut Td-3). This mark Is tntcted by
cutting ecross the anterior face of the distal tibia
when the leg Is outstretched or stratght. The resull s
a merk across the anterior fece of tha libia (Td-3) that
generally conttnues ecross the faca of the lateral
maleolus. final1y intersecting the calcaneus (Figure
4.26, a and b. Tp-t]. If this movement s somewhet
lower, the marks may tntersect the anterior face of
the estregalus at essentially midpoint, as shown in
Figure 4.27e. In such a silualion the knife may "turn
the comer" and mark the medial face of the as-
tragalus (Figure 4.271) al the eeme time rnerkmg the
medial side of Ihe distal tuberositv of the libia (fig-
ure 4.26e). Rarely there may be ~ prominent mark
across the articulator surface ofthe distal tibia on the
smooth promineot ridges that caotact the astralagus.
Thisresults from a kind of swiping cut as the joint is
disarticulaled after the euts just described have been
The dismemherment of the tibia-tarsa} joiot was
one of the aoatomical areas given special attention
by Henri Marlin in his seminal study of butchering
marks. He supplies delailed descriplions of marks
observed io Ihe Mouslerjan fauna as well as accurale
reconstructions of the dismemberment procedure
(Henri Martin 1907-1910:251-274).
IotlfOl vi..
mediol vitw
omlflot vi,.
Oiltol ri9ht,
ventral vie.
~ .... T'-2
onterior view
PrOllimol rifiJht,
ventral vie.
mldio) vi,.
J>!oollimal rifiJht,
lateral view
FlGIJRF. 4.26. Mol'ks inflk'ed UIJ 'he tibia dUl'illg dismemberment.
.. ,"
120 4. Human Mcdes o/ Bcne Modificaliol1 Cut Morks: Their Formond Plccementon Specific Bonos 121
M.dlol vi
Llfl diJIOI mllotorJol,
on1160' vi.w
Lallrol y.w
I C]
Le" colCClneus,
medial vie.
~ 7 TC-' ---1\
Left coltoneus,
dorsal vi...
... "TC-I -1-.
Lel! CakO(lfHa,
lot,rol vte...
nGURE ".Z7 Morks produced on 'he 'arso/! ond th'e me'a'arsol during dlsmembernrenl.
re-t TA-2
~ ~
.:. '
~ . ,
, ~ '
Lttt Q$I'oQolus, LeU astrogolus, Le" astrQ9(l1us.
lateral vi,... onteior vi, .. m'diol vi,...
(d 1 (.1 (O
Venlrol vil"
glencid cevty are mostlikely lo be seen in localions
of consumptton. unless there Is prccesstng for dry-
ing, or in sttuatons where relatively large entmals
are being butchered and the parts are desttned for
trensport rather than processng:
As in the case of cut marks en the scapula. few
marks on the proximal humerus (Figure 4.30) can be
releted to initial butcbenng. When rnerks do occur
they are coneentrated in basically Iwo locettons and
occasionally in a thtrd. Mosl eommon are short
marks on 'he lower "llp" of !he condyle [Hp-I}, as
seen in the postertor aspecr. The next rnost frequenl
is a mark on the apex of the lateral luberosily com
~ T d - 3
FIGURE 4.27. (continued)
Primary bulcheting traces 00 !he scepula {Figure
4.29) seem to be restrictad lo merks that encircle the
glenoid cevty. There is a tendency for Ihe marks to
be00 Ihe lateral tace and lo exhbit sorne concentra-
Iion at the cngn of the trceps brachia. Analogous
cuts may orr-ur up the scepule around the neck.
Dsrnemberment al ths joinl s most oommonly a
secondary butchering operation. unless the animal ls
quite large. This means that considerable numbers of
innided marks 00 Ihe scapula in Ihe area around the
be expected lo have been inflieled durtng secondary
butchenng and processtng for rnarrow rather then
durng primary dtsmernberment.
ducad when a knife is inserted directly into the jclnt
(see Figure 4.12). II ls my mpreseton that this
rnethod of dserticulation is more common with
metal lools and that marks in other localions can he
expected if stone tools were being used. Most Jikely,
euts with stone tools should be coneenlrafed 00 the
anterior 115 well as Ihe lateral and medial faces jusi
ebove tha eprcondyles. In mosl cases such marks can
expected on the lateral (MTp-3), anterior (MTp-l),
and medial (MTp-2) faces. These rnay be produced
during primary and or secondary butchenng.
00 the distal end of the metatersal. marks relang
to dismembermenl can be expected acreas the vent-
ral face of the enndyle. as is shown in Figure 4.27
Sorne nicks mey occur along Ihe maegtns of the in-
tercondyiar channel (MTd-3). Such marks are pro-

l' )
Right di'lol,
medial vi.....

Right prolimol,
lalDrol vi, ....
H'-2 (-::.. ..J8
Lolero! r",berosily
Right dislal,
venlrol vi'"
RiqM prQ.imol,
pOlle'ior vie....
R;ghl distal,
cnterier vie.
FIGURE 4.30. Marlcs produced on the hume"" during dlsmembermenl.
Righl proaimal,
meltial vie....
Right dislal,
lolerol vre....

"1 S-Z

, ".'
" ..


4. Human Modas of Bone Modifka!hm
\ '
proximal end of the humerus with choppers seems a
strange wasle of time and effort when two simple
cuts and a [ittle leverage are all Ihal are needed lo
dlajoint Ihe bones in queston.
If the proximal humeros exhibits rlatively few
cut merks. probably beceuse of the ease wtth which
FIGURE 4.29. Marks produced 0fI the Kapula durlns
"l=S)( TE-t

Anterior vilw ot rill"t tar.u.
FIGURE 4.Z8. The arficulatlon 01 the forsol, with lbe
monly located 00 Iba posterior Iece. The merks 00
Ihe lateral tuberosty and tbe marks 00 the scepula al
theanterior rnargin o the glenoid cavity are believed
la be made al the sama time by a simple cut orlgtnat-
tng 00 the anterior sirle o the jotnt wilh the joint
artieuJaled in a farly stretght fashon. A transversa
cut is simply made into the ont frcm the front. re-
sultfng in the knlfe impacting bone 00 the lateral
tuberoety and along the anterior lip of the margio lo
the glenod cevity. Afiar the jont te cut into from Ihe
fronl, an anelogous cut is made from the rear. result-
ng in marks on the lip of the humeral condyle
[Hp-f] as well as along Ihe posterior margin cf the
head oC the scapula. These cuts can be made very
quickly. The dismemberment of ths [ont is rete-
vely easy. since the condyle ts large and the jonl ts
rether flexible. Leverege s used to greet edventege in
seperettng the scapula from the proximal humerue.
As wtll be descnbed later, this lotnt s commonly
sepereted from Ihe cercase prior to the removel al the
scapula wben large animals are being butchered.
Ths strategy permita the use of the scepula as an
anchor and provdes Ieverage for dislocettng the
jotnt. Argumenta about tbe dfflculty of disjointing
tha scapula from the humeros seem odd to me, as do
proposals ebout tbe use of choppers. Deslroying the

Proximal riQht
Medial wtw
lot.rOI vie.

Int.rior l/l

" \

, ,
il I 1:\1
,': I
, ,

Marks ...hencuttio9
io from Ihe reee 00 "slif'" joiot&

Cubitus-t- I I -tt--Rodius
! /' ,''
, I

, ,
oriented cul marks (see Fi,Rure 4.3of). These oblique
marks may be somewhal higher up on lhe condyle.
almosl at the "ne." Such a placemenl of marks,
coupled with marks on the dorsal ridge of Ihe oJee-
ranon and alons the margins of Ihe olecranon fassa,
4. Humen Modes of Borre Modificalion
Right lalefol I/i,.
FIGURE 4.31. The QrliculotJon berween t1te tllstcd
humeros cmd the pro:rimol rodio-cubilus, showins marles
produced durinS dismembermenl.
the joot may be disartculated, the distal hurnerus
generally sports a consisten! and numerous collec-
ton of cut rnarks. The rncst common mark is 00 nr
8GroSS the medial face 01(he distal hurnerus (Hd-2).
The second most common mark te a short mark 00
the prominenl ridges of the anterior Iace of the ar-
tculator surface (Hd-l). These two cut marks are be-
Iieved lo result from the sama cutung motion, which
also produces the marks along Ihe margn o the
radius (Figure 4.32, RCRp-5). The knfe s run trens-
versely nto the jotnt and turnad inward 8CroSS the
medial face DE Ihe distal bumerus. somettmes even
penetreung lo the medial Iace of the olecranon
{RCCp-3). This is generally done with the leg fairly
well extended or stratght. The second cut commonly
employed in dtsiocattng the humerus from the radio-
cubitus is a diagonal cut mede bestde the lateral
face of the distal humerus. impacting bone 00 the
Iece of the olecrenon jusi besde the semilunar notch
(RCCp-2). This cut makes possible a twisting of the
ertculeton inward toward the body and may also be
accompanied by slight twisting of the lower leg or at
least the shaft ofthc rado-cubnue. This has the effect
of "ectng" the arttculetton out of the socket obique-
Iy toward the body. Once thts level of disJocation ts
echeved il is a simple matter to cut Ihe remaining
strings of ccnnecttve tissue and seprate the bones.
Ths technlque essumes that {he body is warm end
the jcints are flexible. When the body is stiff andlor
Irozen, a shghtly different method may be employed.
The first task. making the lolm flexible, is accom-
plehed by cuttng down from the rear between the
olecrenon end posterior surface of the distal tibia.
Thls cut ts Irequently essociated wtth an attempt to
flex the otnt as much as possible at Ibe sorne time
the cuts are bemg made, resulttng in a series o short
cut rnarks along the dorsal crest of the olecrenon and
elong the margins o the olecrannn fosse. which is. of
course. on the posterior face of the distal humerus
(see Figure 4.31). Once these cuts are made and the
joint Is flexed. butchenng may proceed as befare.
However, I have observed bulchers following Ihis
procedure and Ihere is somelhing of a lendencv lo
make tbe cuts across the anterior face jusi above Ihe
articulalor condyles of Ihe distal humerus ralher
Ihan below them. as is the case with warm animals.
When Ihis is done Ihe leg is sil!! nexed, and if Ihe
"comer is lurned" Ihe knife runs across the medial
face of Ihe dislal condyle. resulting in obliquely
FfGL'RE 4.32. Mario; pmduced 00 Ihe rodio-eubitu, durioS dismembermenl.
Fll;IIKt: 4.:1:1. HU,'I..I j,'","r I/w ml'al"l'u",,,,J (r.. m 11", ,"ur
betrays the butchering of ths [olnt when lt is flexed,
which is most common when the animal is stiff.
Sliffness may be reiated lo Ihe dismemberment being
part of secondary butcherlng acttvtttes. or lo
scheduling problema al mass kills. In this regard it is
interesting that the marks lustreted from the
Eden-Farson sile (Frison 1971:264, Figure 3a-d] all
exhblt ratber oblique cut marks high up near the
neck of the distal humeros. Earlier in this chapter 1
potnted lo other evidence that seems lo indicale the
butchering of stiff or frozen antmals al that stte. It is
equally interesting that the "stff jont" patlern of
butchery seems indicated al the Jurgens site (Wheal
1979:64, and Figure 31). I would guess tha! secon-
dary bUlchering took place. wilh disjointing occur-
ring aner meal Was filleted from essenliaJly complete
front legs. Disjoinling was most likely relaled lo mar-
row procuremenl after the joinls were both stiff and
Most of the (;ut marks occurring on the proximal
end of Ihe radio-cubitus (Figure 4.32) have been dis-
cussed in Ihe comse of considering the distal
The distal radio-cubitus frequently exhibits
marks across its anlerior face, right along the edge of
the arliculalor surface, Alternalvely, marks may
occur only on Ihe slyloid praccss, indicaling Ihat Ihe
cut was made lower down and lllans Ihe lateral face,
impacllnR the carpals ralhl'lr than the edge of the
distal raoio-cubitlls, The lalter marks are more com-
mono On the inferior face of the articulator surface
Ihere may be transverse marks on the prominenl
ridges, These derive from Ihe insertion of the knife
into the joint afler il is partially dislocated.
Cul marks may occur on any of Ihe exposen faces
of Ihe various carpal bones bul are mllch more com-
mon along lateral and medial wilh far fewer
occurring on Ihe anterior ano posterior laces. George
Frison (1970:12j describes the of marks nn
the hison carpals from lhe Clenrock site, a very gaael
description uf buldwril1g of Ihis joinl: "Of lhe ulnar
carpills and ffH]al cilrllals. 4:1% of Ihe fOfnwr anrl
4. Human of Bnne Modjfjmrion
39% of the latter bear cut marks. This is slill a corn.
mon method of butchering in Ihe field and after cut-
tng in this rnanner . . the foot muy be snepped off
[Frisen 1970:121." Whal Frison ls descrtbing ls a
simple patr of cuts. une across the arlic.ulalion on Ihe
inade marktng the medial of "radial carpals." and
another across tha lateral fece marking the "ulnar
carpals" and nol infrequently the slyloid process of
the dislal redo-cubttus. Once such cuts are made Ihe
otnt is "snapped" in half by pushing in or pullng
oul on the joint. as is tllustrated in Figure 4.13.
Essenlially the .'lame marks noted for Ihe metatar
sal can be expecled on the metacarpaL
I have rarely observed marks on Ihe phalanges.
This experience is held in common with most of Ihe
other North American researchers observing Ihe
properties of cul marks. On Ihe other hand, Henri
Marlin reporled numerous cul
phalanges from Ihe Mousterian sile of La Quina. I
have suggesled earlier Ihat these cuts derive primar-
ily from skinning when pains are being laken lo skin
out the fool in greal detail. This is done among the
modern Eskimo only for oblaining skins from which
soeks and shoes are manufactured. This is almosl
exclusively a faHand early winler aetivily among !he
Nunamiut Eskimo.
FiIle!ing Marks
Slrie!lv speaking. the marks previonsly described
for Ihe Ihorar;ic and lumbar verlebrae associaled wilh
Ihe removal of the lenderloin are derived from fillel-
ing (Figure 4.211. Onl' rnight al so argue Ihalthe re
moval of he tongue (Figure 41lR) is ao acl of fillel
ing, Bolh of Ihese acliolls lIre eommon]y eonducled
as parl of primflry bUlt;hcring Filleling for
lransport as by Binfonl
describe!! lIlllong tlw lJolm !KUll,& ,IYellen 197"".
2791. amI as amung oul
mass kills for purpOSt'Snt RPrH'raling lar.w quanlitie5
of lnl'ar (as was prohabtv lhp silua!ioll al the

Cut Mcrks: Their Form and Ptocemenr on Speclfic sones
tsen-Cbubbuck site-WheaI1972) are all Instances
of primary butchertng aclivilies. We might think of
primary hutchering actvtties as thcse that are timad
or scheduled with respect to the killing activtea.
whereas secondary butohenng andror proeessing ac-
tivilies are timad or scheduled with respect to
achieving (he goals of plactng meat inlo storage
andlor dislributing and preparing it for consump-
tion. Secondery butcberng normally consiste of fur-
ther segmentatton of the parts generated al the lime
of primary butchering, and/or lletlng prtmartly ap-
pendicular parta. 1 have never heard or seen
documenled filleting aclivities connucled with re-
gard to any axial skeletal parls exceptlhe lower lum-
bar vertehrae and pelvis. Thus we can generally ex-
pecl an overlay of marks deriving from filleting lo be
reslricled lo Ihe lumbar verlebrae. lhe pelvis, and the
leg bones. Even in those cases where filleling is con-
ducted at locattons of primary butchermg. as al a kili
ste. t is a second-stage activity; animal s are field
butchered into basc analomical segrnents. which are
then filleted. This means that the charecter of the
segments condtttons to sorne extent degree lo
which fiHeting marks will overlay or , alternahvely,
be the only marks remaining from butchertng on cer-
taln parts.
1have observed numerous ects of lleng carrted
oul by the Nunamiul Eskimo and the firsl mpresston
is that tt is done very quickly and with seemingly
Hule efforl. For instance, Figure 4.33 shows !he
bones of Ihe rear leg lying beside Ihe meat removed
during fiHeling. This filleling was done by bulcher
in lhe field so rapidly thal 1 did not have time lo
adjusl my camera and take a picture before he
4. lIumun Modes of Bone ModifiC(ltion
Cut Mnrks. Their Form nnd P/ocemen! on S.It'J'k Eones
H(jlIRE 4.;14, fhmt after uf
nf remonng al a praressiUR ocotian aftcr fille/iug.
sborter and are rarely exclusively longitudinal. as ls
shown in Figures 4.06 and 4.36, Long, Iongiludi.
nallv orienled cuts are made bUI these tend lo be
mde along the sides of Ilw bont' relative lo Ihe
orientalion of the part during filleling operalions.
Musl commonly, Ihe legs resl nat dming filleling
operalions, eilher on Ihe lateral or medial "sides"
rclalive lo lhe living animal. This means Ihal cllls
will appear along Ihe anterior anri poslerior faces of
Ihe hone when viewed in analomir.al orienfation.
Knives are fIln along fhe "edge" of Ihe bone anri do
not, slrictly speaking. produr.e cul mrks; irregular
longitudinal sha!low sCfillches may be prorluced. but
an relalivelv rare. Mu<:h more common are shorl cul
mlrks freqlHmllv ubliquely lo lhe longiludinal
spl'd of lhe bolle anri cOllcenlri11t'd nn uolh thp an-
terior ami !;l('es. Sur.h tl\:trks will be clus-
along tha bone and rolltng back the meal so lhe bone
tan be pulled free frnrn the rneat. There are generally
Iwo types of cuts: la} inilial long: longiludinally
orienled. bone-exposing cuts, and lb) shorler, more
oblique [uls mad{l lo the unrierside uf lhe exposed
bone to free il frum Ihe mass of meal amilor sever
muscle insertions. Gi\'l'f\ such a slrlegy, il is nol
lo see orienled cul marks
anri llhorler. more obliquely orienlfJd marks along Ihe
poslerior Uf <.Inlerior surfaces of 10nR-bolle diaph-
yses. Goon P-Xilll1ples of long, longitlldinally oricnled
filleting marks are seen on Ihe scaplllll in FiRure
4.06. Similar marks are shown on Ihe pelvis illl1s-
lraleri in Figure 4.3fi, Thp fillcling marks indir:aled
in Fifi\lIre 4.:U; wp-re (Jn knowl1 filleter!
SpucinlPns mong lIw NII!l'lflliul
Fillf'1 ing (l1rb (lll bOllPS nf 1111' are much
Deplcted here are the parts that were Wleled (or
drying from 11 carihou killed earler (see L. R, Bn-
ford 11978b:223-2281 for a descrlption of ths epi-
sede}. 1 am strongly suggesting that mere wil1 he
dtagnosttc pattems of articulation, panems of part
association. and spatial features thal will covary
wifh disf inctive palterns of mfhcted marks when
illetng is a mejor cr dorninaut activily al asile
Ths is an argumenl to bedeveloped at another lime,
since my majar concam here is to describe the marks
thet dervu from Iilleting.
The rst fact lo emphasize is Ihal lleung marks
are ulmost exclusively longttudtnellv orienled with
respect to the bones on which they appear. The very
act of filleting dlctutes this patlern. stnce Ihe butcher
ts essentally rernoving the bcne from the mess of
meat around il and Ihis e best achieved by cutting
finished me ktng his cuts and rarnoved the llets.
The situation Hlustrated is rather rnrnmon-c-the but-
cher is aware lha\ lletng will be the next step. and
no transport problerns extst {e.g., llettng is to be
done in (he seme place). Segmenting of the legs is
rarely done during priruary butchering; tha! ts, the
lower Iegs are rarely removed separately. Instead
Ihe l,mtire leg is removed and it ts Ihen lleted and
disposed of, a complete or nearly complete leg re-
rnaining lergely arriculated. This stuatton is Wl!\1
ttlustreted in Figure 4.34, whare arttculated front
legs mtnus the scapule (see L. R. Binford l1978b:l 00]
ter a dtscusston of the treatment of the scapula by
the Nunamiutl are shown jusI alter having been
Filleted. The ccnsequences of ths behavtor at pro-
cesstng locations is frequently apile of nearly com-
plete articulated legs. as ts shown in Figure 4.35.

. ...
4. Human of Bnae Modification
Cut Marks: Their Formcnd Plccemenj on Specjic nones
Lelt onf,rior vi, ..
lo exhib mosl such culs is the redo-cubttus. and as
s e1ear thts is the bone where the articulation wif h
the humerus is characterized by the rnost irregular
and crevtce-Ika areas. making filletiog very dif-
ficult. Figure 4.39 shows the Iocatton of marks In-
flieted on Ircnt leg bones during known sttuattons of
meet filleling; the patlern s generally the same on all
the bones. The rnarks tend lo be very superficial,
rarely deep: 00 freshly fiJIeted bones they are only
visible as cuts through the perioeteum lhat did no!
penetrate the bone. Unfortunately, 1did not relurn lo
Left j)ollerit>rvi, ..
Figure 4.38 illuslrales the distal ends of rear leg
bones with merks derivad from filleting operations.
Avetv diagnostic patlern. one 1call short chevrons.
is cleaely indicated. Thesa are pared clusters of
short paral/el cuts obliquely oriented and originating
from opposlte sides of the bone. These are generally
more ccmmon on the distal ends of bones where
tbere are muscle Inserttons. as in the case of the dis-
tal tibia and perticulerly the distal radto-cubttua.
Theyare less common on the bones of the upper leg
such as the humerus and femur. The bone that tends

l. I P5-6
FIGURE 4.36. Morks produced 011 the pelvis doring fllletJns
fiGURE 4.37. Mark, produced on 'he proJ(imal 01 upper .".ur leg bones during [iIleUng.
Right m'diol vi...
Tp-4 \;
T,, 't,
Righl anterior vi, ..
R;9ht holf,
dorlol vi,,,,
marks concentreted in tossae oc around indentalons
in the bone such as the area along the lateral sde of
Ihe tibie! crest. Figure 4.37 illustrates marks known
lo have been produced on the proximal eods of
upper rear leg bones during fllleting by Nunamiut
Eskimo.Jt should be cleer that Ihe marks are (o) gen-
erally obltque. lb) generally located on the "neck'' of
tha bone. and (e) cornmonly in recessed pteces
where one would have to cul the meal out ralher
than slrip it behind a smoolh cut.
Righf hall,
untrol Yi...
tered where the shape o the bone is irregular and
where there are numerous muscle nserttons. lrregu-
larity in shape is most eommon in the lmmediete
area of the artculetcr ends. hence we can expect a
numbee of short cut marks in the area of the "neck"
between the epiphysis proper and the linear shaft of
the dtapbysts. As anyone who has carved mea!
knows, it Is where bones are irregular in shape that
there is Ihe mosl problem Coc Ihe cerVet. This is
equally true Coc filleling. hence we can expecl many
4. Human nfDone Modifir.nrioll
rillht ,odio-cubihll
RCp-7 RCp-6

Loterol vil"
'ioh! hum.ruI
MlidilJI yiew
Loterol vi,w
Distolleft mllol(l'501
Distal riOM 'emur
Posterior vi....
Distolleft tibio
vii" M.diol vil"


Lorero! vilw
0;"01 left rodiocubi'u,

Anterior vil'" Mediol vil"
O;IIGI ri",tll humerU$
Anterior view

Anterior ve ... A,,,,ter;or view
FIGURE 4.38. Marks produced on Ihe distol eods 01mar les Inmes durirtlJ fillefinR
FIGURE 4.39. Marks produced on {ront les bODes durioJlilletlng.
Jir.. '
Albuquerque a total assemblage of bones from a
known fillefing operaon. so 1 cannot provide reh-
able nformaton 00 Ihe frequency of such marks.
However. it was my impression that they were pro-
duced 00 a relettvely small number of the bones ac-
tually {ilIeted end that Ihe frequency increased as Ihe
limbs belng fil1eted were supercally drted. rnaklng
the "thin" ends such as the distal rado-cubtus and
the distal tibia tougher and more dffcult to pene-
trate for startlng the filleting operaton.
The literatura is relenvely sent on these types
of marks. and when they are descrtbed they are rarely
recogntzed Ior their speclfc signftcance. Rather.
they are rnost often lumped with dismemberment
marks as "butchering marks." An exceptton in this
regard was Henri Martin {1907-1910J, who called
them "rnarks of evisceration" and correctly iden-
ted sorne of the short oblque "chevrons" occur-
ring on long-bone fregments as being produced duro
tng the removel of the meat (see Martn 1907-1910:
Plate LX, No. 2, and Plate Nos. 2, 3, and 6). No such
fiUeting marks were reported Irom the- Eschelman
stte (Gllilday el al. 1962). Smarly. Frisan (1970)
cited no such marks from the Glenrock stte, nor were
they noted (Frisan 1971) et the Eden-Farson stte. 1
observed only ene example in the Mousterian mate-
rtals from Combe Orenel-.-e distal rado-cubltus from
Level K (Quina Mousterian). On the other hand,
these marks have been described from the Jurgens
site (Wheat 1979: Figures 31. 32), where they were
lumped with marks of dismemberment, breakage,
and other characteristics that the author thought re-
lated to butchering.
Marks Produced
during the Prepafation of
Ports for Consumption
One of the most common methods of preparing
parts for consumption is in fact filleting, particularly
when the meat is to be bailed or used in a stew. This
introduces sorne ambiguity into Ihe significance of
fiUeting marks. This ambiguity is fnrther exacer-
bated by the fact that when leg bones are prepared
for marrow cracking they are typically "cleaned,"
which consists of cutting off adhering secHons of
meat or tendon that would modify the way the bone
4. Humun Mudes a! Hone Modifica/ion
would break when impacted with a percusston blow.
Figure 4.40 shown lean Rulland cleaning marrow
bones prior lo their bcmg cracked for marrow. with a
woman's knife or ulu. of merrow
bcnesproduces shcrt parallel cutaieeks essenrelly
Indtsttngutsheblefrorrrthe ftHefhig-h...-erb i1lustrated
in Figures 4.37-4.39. However, much attention s
gtven to the metapodals, distal rado-cubttus, and
distal libia, all bones thal yield ether no rneat or very
Httle. This rneans that "llettng marks" may well be
abundant en the metepodtals. bones not eesly con-
ceived of as having been lleted! It ts interesting that
loe Ben Wheat observed this at the [urgens stte. and
commented as follows:
Ltght diagonal cuts were found on a number of bones
including rnetapodals. 1I is not Iikely that Ihe cuts on
rnetapodlals were assoctared with filleling meet. snce
there is very Hule maat on such bones. so perhaps sorne
olher reeson must be found for these cut marks [Wheat
Cleaning the bones of segmente of meat and ten-
don is nol aU that is done prior to cracking. Toe
is the
remcvalot the-penosteure-rrtbe-eeee-ee-be JI _.d.
The Nunamlut invariably do this by scraping it back
with the edge of a knife. a rough surface on a ham-
merstone, or almost any handy crude scraping too1.
This means fhat lonsitudinal sg8tch8 and striatio/p
(Figure 4.41) along the shaEls of long banes are
commonly produced when bones are prepared for
cracking during marrow processing. Such marks are
noted in Mousterian assemblages. My notes record
18 examples (striated bones) from the site of
Grenal, but the fragments of long-bone shaft were not
saved at this site, so the parts where most such siria
tions would be expeeted were in fact never systemat.
ically examined. Henri Martin (1907-1910: Figure
LXII, Nos. 1, 3, and 5) reporls examples from La
Quina. Most analysts do not examine aH the long
bone splinters for such marks, hence there is Uttle
mention of this type of mark in the literature.
Modifications other than those jusI described may
occur as a result ofvarious forms of food preparalion.
Howa part is trealed (filleted or not), what size it is
reduced lo befare consumplion, and how it is re-
duced, [chopped, cut. broken, etc.) are all variable
strateges that relate lo the methods of food prepara-
Cut Morks: Their Form ond Plncemeru on Specifk BOlles
FIGURE 4.40. Jean Rullond cleaRing malTOw bones lar cracking.
FIGURE 4.41. Slriations on bone
believed lo hove been produced duro
ing removal 01 Ihe periOSleum. On
top 01 Ihe sfriotJons is pllting be-
Jieved lo have been produce(1 by
gnawing animals, (Found in rhe
Mouslerian site 01Combe Grenal by
F. 8fJrdes.)
13. 4. Human Modes nf Bone Modljtcution
Cut Markli: Their Form ond Ptocemenr on Specifit; senes
_O, _._. _. _
tool-ustng man trom Ihe carnivorous actlons of cther
anmals. Traces referable lo sktnning: dismember-
ment, filleling for transpon. and marrow consump-
tion are al! perhaps relevant. It ts hoped that enough
has been demonstrated regarding pauemtng in cut
marks lo encourage others lo describe tbetr material,
so as to begn the tesk of developtng a Iarger corpus
o comparative malerial for study and use in tbe furo
ther speccetton of dtagnostic cherectensttcs reli-
ably relerabie tu specifc actions carried out in the
past. {See Table 4.04.)
on. For nstance. whether something s bmled.
roested. nr eaten raw, and the types D contatners
used in food preperatton, as well as the methods
used (stone boilng versus dtrect boiiing, etc.] can be
expected to iecve subtJe traces diagnostic nf eccb.
Researching these dfferences is no! the subject o
thte dscusson: retber.J have been concerned with
presenttng descrtpnve tnformaton regarding the
traces thet tool-using man might beexpected lo have
left on bones when he hunted and killed. This in-
formation might serve lo distinguish the acttons o
TABLE 4.04
l n ~ e n t o ' Y o/ Described Sklnnlng afta' Bllrcherins Marks
Codo Part end Activity pro-
number" description ducing mark
s-i Transversa cul on occipital con- Ihsrnembering
S-2 skull spllt inlo left and rlgbt halves Food preperation
S-' Transverse chops acrees crentum Dismembering,
ebove end below the antlers or coneumpton
So< Cuts circ1ing the Mse of antlers Skinning
SO, Hole pounded into the fronlal llrea Killing,
food preparalon
Cut on maxilla just above third Dismembering
S-7 Nose cut off Consumption
M-1 Transverse cul on Ihe inferior sur- Skinning
face of Ihe mandibular symphasis
M-2 Marks across the masseteric fossa Dismembering
M-' Diagonal cul on medial surface 00- Dismembering
low P-3 and P-4 [lon8 ue)
M-' Diagonal cut on Ihe laleral face be- Dismembering
hind the third molar
M-s Transvarse cut on the inferior sur Dismemberinll
face of the mandibular condyle
M-6 Longitudinal cul across masseleric Dismemberinll
Guilday el 01. 1962; Figure 2
Guilday el al. 1962; Figure 7
Martin 1907-1910' Figure XLIII.
Nos. 9. 10
Frison 1970; 24
Frisan 1971: 263; Guilday el al.
1962: Figure 2; Pannale-e 1965:
Figure 1
Guildey el al. 1962: 65
Guilday el aj. 1962; Figure 2
Parmalee 1965: Figure t; David
1972: 317
Wheal 1979: Figure 33b
Wheat 1979: Figure 3c
Martin 1907-1910: Figure XLVIII,
Nos. ]-:.
Wheal 1C79: Figure 33c
TABLE ".04-Conlinued
Code Part and Actlvity pro- Figure
number" descrrptton dudng mark number R6farenGtlli'
Cervrcn! venebme
CV1 Transversa cuts ecross the proximal Dlsmemharng 4.20 Guilday el al. 1962: Figure 2; Mar-
ventral surface of atlas
tin 1907-1910; Plata XLVII. Nos,
3,4. 5
CV-2 'rrensverse cuts ecross posterior Dismembering 4.20 Martin 1907-1910: Plale XLVlI,
ventral surface o alias (stiff body)
No. 6; Parmalee 1965: Figure 1
CV.J Transversa cut ecrcss anterior Dismembering
No' Frisen 1971: Figure aee-dd
ventral surface ot axis (stiff body) shown Parmalee 1965; Figure 1
CV-' Longitudinal spltttng of vertebrae
CY5 Transversa chopping of axis Dismembering
CV-6 Transverse cutting or chopping of Oismembering
No' Martn 1907-1910: Plete XLVI, No.
aixth vertebrae
shown 7,64
Thomcic vertebme
TV-, Transversa chop or cut merks along Dtsmembertng
dorsal spine intn or between cen-
Ira of eeccnd and Ihird vertebras
TV-' Longitudinal cut along base and Filleling 4.21 Frisan 1970: 20-21, Figure 14; GiI-
lower part of the dorsal spine tenderloin bert 1969: 290; wheat 1979: 67
TV-' Transverse cut ecross Ihe inferior Dismembering No! Guilday el al. 1962; Wheal 1979:
surface of the cenlrum (ribs) shown Figure 33,11
TV' TranS\'erse chop or cut marks along Dismembering
dorsal spine inlo or belween cen-
tra of thirteenth lo fourteenth
TV-5 Longiludinal marh iusl below Ihe Dismembering 4.21 Wheat 1979: Figure 33,11
articulalor processes (ribs)
Rlbs ond slernum
RS-l Marks along superior surface just Filleling No! Marfin 1907-1910; Plala XLII, No.
lateral lo the rib head (tenderloin) shown 5
RS-2 Dislal ends of Ihf' rlbs cut off Dismembering No'
(ribs from shown
RS-J Transverse cul on ventral rib Dismembering No' Frison 1971: 265
surface jusI lo Ihe side of Ihe [ribs) shown
rib head
RS4 Marks on Ihe venlral surface of Filleting No'
slemum plales shown
Pelvis nnd socrum
Sat:r\lm Irirnmed down laleral edges Secanoary No!
buh:hering shown
4. Human Modas of Bcne Modificotion
Cut Morks: Thetr Fnrm ond Plocemenl on Speclfic Bones
ode Part and Activity pro- Figure
number' descriplion duelng mark number
PS-Z Marks msrde the acetabulum Dismembering No'
PS-3 Marks serosa the lateral tace FilIeting 4.36
or tbe pubis
PS-4 Ischtattc apine Irimmad off
PS-5 Cut or chop Ihruugh. the pubic Dismembering 4.36
PS-6 Cut 8CtoSS deep fossa in fronl of Filleting 4.36
PS-7 Marks ebove the acetabulum on arm Dismembering 4.22
o Iltum
PS-8 Marks below acetabulum on arm o Dismembering 4.22
lsch.ium 4.24
PS-9 Marks ctrcltng the rim of the Dlsmembering 4.22
PS-I0 Marks below ecerabulom en arm o Dismembering 4,20,4.22.
pelvis 4,24
Fp-l Marks on the neck o the femur Dlsrnemberlng 4.25
Fp-2 Marks on the hall of femur head Dlsmembarlng 4.25
Fp-3 Marks cfrcling the margln of rhe D1smembering 4,25
fernur head
Fp-4 Transversa marks on lesser tro- Filleling 4.25
Fp-5 Transversa merks on lateral surface Dismembering 4.25
of greetee trochanter
Fp-6 Short marks in the fossa elong tbe FiIleting 4.37
neck of the femur
Fp-7 Short rnarks on the neck of greater Filleling 4.37
trocnanler, anlltrior face
Fp-8 Short "chevron" marks, both !rans- FilJatng 4.37
verse and oblique. below the necl
of lhe femur both 1leral and me-
dllll faces
Fp-9 Longitudinal mark on upper shaft of fJlleHos 4.37
Fd-1 Transverse aeross posterior surface Dismembering 4.25
jusI above condyles
Fd-2 Nick marls on upper marglns of D1smemberi nll 4.25
Ihe patellar surflllce--Irochlea
Fd-3 I.ongiturlinal mar\:.alollS lhe lateral Dismembering 4,25
face nf Ihe medial condyle
Gullday et al. 1962; Figure 2
David 1972: 317
Guilday el 01. 1962: Figure 7
Guilay et 01. 1962, Figure 2
Wheat 1979: Figure 32
Martn 1907-1910: Plate L. No. 12;
Frison 1970: Figure 8a-b
Martn 1907-1910: PIMe L. Nos. 8,
9,10,11; Guilday el 01. 1962:
Figure 2
Wheal 1979; Figure 33
Wheat 1979: Figure 33
David 1972: 319
Wheat 1979: Figure 33
Guilday el o/. 1962: Figure 2
Code Part end Activity pro- Figure
number" description ducing merk number References-
Fd-4 Short "chevron" marks obltquely Fillettng 4.38 Wheat 1979: Figure 33
Fd-5 Shcrt oblique merks on the anterior Filletlng 4.38
face above petellar surface
Tp-1 Marks on or around inlercondylar Ifismembanng 4.26 Guilday el oJ. 1962: Figure 2
tuberclea-c-partir-uiarly the lateral
Tp-2 Mark ecross the posterior ece of Disrnembering 4.26
the Iateral and medial coodrtes
Tp-3 Shorl oblique rnarks un the lateral Filleting 4,37 Wheat 1979: Figure 33
tace of tbe tibial crest
Tp-4 Obhque marks on medial tace just FiJleting 4.37
below the arliculator suriace
Tp-5 Longitudinal merks 011 the upper FiIleting 4.37
tibial shaft
Td-1 T'ransverse merks acrees medial Drsmembering 4.26 Guilday et 01. 1962: Figures 2, 7
malleolus and ust ebove on distal
Td-2 Transversa mark ecruss Inferior Dismembering 4.26
asttculetor surace
Td-3 Teensverse cut ecross anterior Dlsmemberng 4.26
face between the dorsal peojectlcn
and the medial malleolus
Td-4 Short "chevmns" on the anterior Filleting 438 wheet 1979: Figure 33
face of the distal libia
'rcncts Martn 1907-1910: 231-273
TA-1 (Astraglllus) transverse mark at Dismembering 4.27 Guilday el al. 1962: Figure 7
margina of the anterior fece. Frisan ]970: Figure 8c, d. e
midway on Ihe bone
[Astragalus] marks across medial Dismembering 4.27
face midway on the bOlle
TC-t (f'.aicalleusj marks along he proxi- Dism",mbering 4,27 Guilday el 01. 1962: Figure 7;
mal margin of Ihe lateral face Frisan 1970: Figure 8
TC-2 Marks llcross dislal end lhlber clll- Dismemhering, No!
r.is) filleting snown
TC-3 Marks 011 dorsal chest midway be Filleling. 4,27
lween luber calcis 8nd thearticu- hanging
lator surface carcasa
TE, (Eclocunl!\form) lmnsverse mark Dismembering 4,28
along m a q ~ i n rf anterior face
TIIle-! (Nllvlculllr-clIhnid) trllnSVf'rse Oismemharing 4,28
m a r ~ al:ross laleral and or lIuteri-
or fllt;e
~ - ~
4. Humen Modes of Bone Modifkol()l1
Cut Morks: Thelr Form ond P!ocemen/ en Specific Bones 141
TABLE 4.04--Conlinued
TARLE 4.04-Conl;nued
Cod. Pan and Activity pro- Figure
description ducing rnark number Rl:lferences" code Part and Activy pro- Figure
number" desortptton ducing mark number Referenees'
MTp-l Transversa mark acron anterior Dismembering 4.28 wheer 1979: figure 33
Hd-J Marks 00 the upper margms of 'he Dismembering 4.30, Whea11979: Figure 31: David
face o proximal condyle
elecranon fossa (stiff body) 4.31 1972; 317
MTp-2 Trensverse mllrk 00 the margn oC Dismembering No'
Marks 00 the ventral edge o the Dismembering 4.30 Whellt 1979: Figure 31; David
medial fece
lateral condyle 1912:3\7
MTp-3 'rransverse mark on the margin of Dlamembering 4.28
Hd5 Transversa mark ecroes anterior Dismembering 4.30 Marttn 1901'--1910: Plale XLIX, Nos.
lateral face
face above ertcutator end 5
MTd-l Traneveese merk across posterior Dismembering 4.27 Wheat 1979: Figure 33
Hd- Oblique sheet "chevrcn" marka Filleting 4,39 Wheat 1979: Figure 31
fece of both condyles
c1uslered on neck of distal end
MTd-2 Circular mark around the distal Skinning 4.16 Martin 1970-1910: plale LlX, No.
on the anterior face
sha 6: Gullday el al. 1962: Figure 7
Hd-7 Longitudinal mark along medial FiIleting 4.39
MTd-3 Nlcks along mergtns of tntercon. Dismembering 4.21
eres! of shaft
dylar channal
MTd-4 Shcrt "chevron" marks obllquely Filleling 4.38 Wheal 1919: Figure 33
c!usterad on taterel end medial RCp-l Transverse mark inside semilunar Dismembering 4.32
creets. anterioe espect noteh
RCp-2 Diagonal mark ecross lateral sur- Dismembering 4.32 David 1972: 311
S] Marks along inferior border of con- Dismembering 4.29 Martin 1907-1910: Plate XLn, No
face of olecrenon
dyle andlor at arigin of trceps
1, PIole XLIX. Nos. 1, 2, 3; Cull-
RCp-3 Diagonal mar!:. across medial face Dismembeeng 4.32 Guilday et al. 1962: Figure 7;
day el 01. 1962: Figure 2
uf alecranon David 1972: 311
S, Merks along the nack of the scapula Dismembering 4.29 Guilday el al. 1962: Figure 7
RCp-4 Marks on upper margin of medial Dismemberng 4.32 Guilday el 01. 1962: Figure 2
S, Longitudinal marks along hase uf FiIleting 4.06 Frisen 1971: 265; Wheal 1979: Fig-
semilunar notch
spna in both the supra- and in-
ure 31
Transversa merks on anterior mar- Dismemberlng 4.31, Frisan, 1910 Figure 8,f.9; David
frasplnous fossae
gln of radial tubercaues 4.32 1972: Figure 199-2; Wheat 1979:
S, Longltudtnel marks up and down Filletlng No! Whea11919: Figure 31
Figure31; Parmalee 1965: Figure
the medial Iece of the scapula shown
RCp-6 Cluetered oblique "cbevrcn" marks FlIletlng 4.39 Wheal 1979: Figure 31
Humeros belo w lateral and medial tuber-
Hp-l Maeks alang the border of Ihe "Iip' Dlsmembering 4.30 Marlin 1907-1910: Plate XLIX. No
osruee on anterior tace. of radius
of balt. concenlraled on the pos-
4: Gullday et al. 1982: Figure 2 RCp-1 Short longttudnal marks along FiIletlng 4.39 wheat 1979: Figure 31
terior ed8e posterior eresl of Ihe cubtus
Hp-2 Marks on the ape" of the lateral Oismembering 4.30 ParmaleE! 1965: Figure 1 RCd-l Tmnsverse mark across Ihe anterior Dismemberlng 4.32
tuberosity face alon81he arliculalor circum
Hp3 Marks on the laleral far.e of the necio. Dlsmemberlng 4.30 fetenee
jusi below Ihe lateral luberosily RCd2 Tmnsverse rnarks across Ihe 51ylold Dismembering 4.32 Martin 1907-1910: Plate L, Nos.
Hp-4 Short "chevron" mnrks obliquely FilIeting 4.39 process 2.3,4: Guilday el al. 1962: Fig-
oriented along eresl below the ure ,
exlemalluberosity al nserlion RCd-3 Transverse marls on Ihe inferior Dismembering 4.39
of leres minar articulalor surfa;e
Hp-S Sllort "chevron" rnarks obliquely Filleling 4.39 RCd-4 Shorl diagonal "chevron" marks FiIleling 4.39 Wheat 1979: Figure 31
oriented on medial face below on anterior face
Ihe head
Hd-1 Transverse marks across anlerior Dismembering 4.30 Guilday el nI. 1962: Fi8ure 7:
articulator face
David 1972: 317
Hd-2 Transverse marks across Dismemberlng 4.1S,4.30
C] Transverse cut along articulalor Dlsmembering No' Wheal 1979: FigurE! 31
medial surface
margn of "ulna carpal" (os shown
--- pyramldal)
(eoolinlledl ------
4. Humon Modos DI Bnne
Chopping ond Bone Breckcge os Butchering Techmques 143
TABLE 4.04-Contnued
Part end
Transversa mark along anterior
llrtJculalor margfn
Circular cut around distal shaft
Transversa mark across posterior
face of both condyles
Transversa mark across anterior
faca of both condvlee
Short "chevron" marks on lateral
and medial cresls of tba anterior
Activity pro-
duolng mark
Cutldey el 01. 19fZ: Figure 7
Guilday el 01. 1962: Figure 7
""h881 1979: Fgiure JI
Metal tools leave rather obvious marks, 85 Wel
bourne has suggested: hcwever. George Frison
P970, 1971) has suggested that bones were com-
monly used as chopping 10015 in the butchertng of
North American bison. Such implements leave less
obvous marks and. as he has suggested, chopping
results in more distinctive breakage of bones than
any obvious chop marks as referred to in the descrip-
ton of African lron Age metertals.
Figure 4.42 [Frtson et al. 1976:52) nicely sum-
marizas the butchertng-related damage inferred by
the authors from the condtton of the blson bones at
the Hawken site in Wyoming. Twenly-nine different
modifications of bone are indicaled, and oniy 5 are
cut merks: the remaining 24 are sorne form of break-
age believed lo heve been produced during but-
The most extensiva end detetled discussion of
Frison's muscle-strippng argument is perheps con-
tained in his Glenrock (Frison 1970) reporf
Butchering was aecomplished by processes of chop-
ping, crushlng, brsaklng. and cutting muscle origins
IInd tnseruons end then stripping tbe muscles .. the
procesa as described here ts a reconeteucttcn Ircm Ihe
wrtter's interpretation o the evdence of butehering ac-
tivity found en Ihe borres and then Blpplying ths ro the
anatomy of the buffalo along with 11constderaton of the
tool assembfegn [Frisn 1970:12J.
After Ihl! labia WIlI compiled lleamed of an excellent sludy of cutrnerks by Van Den Drlesch and Bceseneck (1975). This work ahould be
ccnsulted .
p .. proximaL d '" dlltal.
. RefurellCflslII"ll only lo lllustr8Uons of mark in question.
FIGURE 4.42. Plocemenl af cul marks on bison bone from tlUt Howkf!fl flite: (1) sowing mads on ventral border 01
manrlibie; (2) t:l1lshinSond breakinA of sacrum: (3) t:l1lminAand bJ"Mkin, of aiecronon; (4--5) choppln, off larerol amI
mediol tuberosilles of humeros; (6) bnmking of scopula: (7) brealdn, o/ ventm} bronche8 of cen-icol.: (8--9) choppln,
mIotroehleo IJndbreoking off mujortrochonteroflemur: (101 breGlein, ofthe pelvl8: (JJ) chopplng o!ftubttr OOJroe: (12)
brealins o/ hvn.llvel'8e processe8 ollumbul'8: (U) bife mGrb from cuttin, baele museles: (J4) breallns ...ln,s of the
ollas; 115)knlfe marb fram cutlin, 100000e lonlue: (16-17) cruming ond brealeing mandlble: (18) brealeinA hyoid bone;
/J9-20, clrappinll ond crushins ro separo'e vertebral column; (21) breaking and chopping IDOSe dorsal spines: (22-23)
choppinS ond breaking of rib8: (24) breollns r:alcaneus; (25) knife marb on culcaneu8 ond asIrtJlltrlus; (26) bntakins
distal end of libia: (27) knife marks on di810lrorliu,-ulna; {2B} breaking off di8!ol rodius-ulna: (29) long oone breakoge
lar marrow. (Reprodueerl with perrnluioll fmm Fri,on et al., J976:52.)
Chopping and Bone Breekage as Butchering
The role of "choppng" and the use of heavy tools
in butcherng are points of "covert'' controversy
among many tnterested in the tools prehtstortc man
used for butchering. Early in the archaeological Iit-
erature il W8S nol uncommonly assumed that large
"core lools" such as hand axes and cleavers were
used in butchery. A most provocative study by J. D.
Clarx and C. V. Haynes (1970) iIlustraled Ihal, in
tenns of actual palteros of association between (0015
and dismembered animal carcasses, Paleolithic
butchering and meat.processing equipment con-
sisted predominantly of small numbers of light-duly
tools, cutting naxes, and small scraping tools, plus
onlya few of Ihe largar elements (J. D, Clark and C. v.
Haynes 1970:409). These findings as well as other
experiences have led me to expect UUle if any majar
"chopping" in mosl butchery. Certainly, Ihis is a
bias on rny parl from having observed so many epi-
sodes of primary butchering among what mighl be
described as the maslers of small-lool butchery, Ihe
Nunamiut. Observations among Ihe Navajo (see L. R.
Binford and l. B. Bertram 1977:91-93) and reports by
colleagues who have worked among conlemporary
African groups affirm Ihe use of the ax. adz, or pango
(a big "bush hife"). indicating that chopping and
the use of heavy tools s a distinct bulchering style,
which may well heve hed counterparts in the
Paleoltthc. This view has been recently set forth
anew (Iones. P. 1980). The foliowing quote by an
analysl of an African Irun Age faunal essemblage may
provide the reeder wlth sorne feeling for the contrasl
between butchering by chopping versus tbe butcher-
ng by cuUing as described from groups like the
Nunamiut (see L. R. Binford 1978b:47--64):
Definite pattems of chopping, tulling and battering
were visible on many bones. These are mosl c1earlyevi
dent on bones of calUeand labras. Sludy of !hese bones
suggested Ihe pllttern of dismemooring of carCllsses. 001"
lined below.
Horns of catlle were chopped from the cranium. skolls
were chopped open for the brains, mandibles wera
snapped off to remove the tongue, hind Iimbs were re
moved by means of a c1ean blow Ihrough Ihe femur
head. ;oints of meat were separated out by means of
chopping blows concentrated al limb articulations. Raw
meal was usually stripped off by mli!8nS of chopping
blows which left shallow chop marks along the lli!ngth
of bones. often concentrated al musde arliculations due
lo Ihe difficully of ehopping lhrough tough tli!ndon.
Bones were eommonly dell for marrow by means of
ehopping blows delivered at the distal end allhough 11
few were Cfllcked. probably with slones [Welbnurne



.. .

lOO 4. Human Modes of Bone Modifico/ion Choppin.'l and BoneBrecknge os ButcheringTechnques lOS
This novel inlerpretation assumes thal cutting is
usad to go through the meal exposing Ihe bone.
However, Ihe butcher Ihan shifls to a hammer and
bludgeons his way Ihrough Ihe bone, achieving a
disarticulation. If correct, Ihis would certainly
resultan! [secnndary] butchaJlng untts lnduded er-
Hculattng proximal and distal ends of long bones and
associated ligaments and tendons, end the long bone
shaft and ils essoeatsdmusculature(meat)... Ihe jomts
were laid asida, if nol discarded, and later bailad to
makesoup or prccessed forany remetmng marrow.The
bone shafland lis meat was cooked as a untt. the maat
overlappingthe bone ends such Iha!the bone was insu-
lated from the fire if roested over an open fire. These
butchering and cocklng pmcesses would resull in the
typesof bone unlts. bonegroups. and breakagapenems
described aboye [Lyman 1976:221.
1 have dfculty wlth this argument. First. many
of the bones ilIustrated in the Clenrock report (Fig-
ure 98, b. e-h, j-o; Figure lOe, h. i , 1: Figure l1a, b. e,
f, , k; Figure 12) are in my axperence identical with
animal-gnawed bones. Second. the patterns of as-
semblege composition given in 'rebles 1 and Z are
consstent with a faunal essembtage thet has suffered
from attrttton: Ior tnstence, the proximal humerus
and the proximal tibia are consistently represented
by fewer examples than are the distal ends. (Sea
Brain [1969J and L. R. Binford end J. B. Bertram
119771 for a diacusston of survtvel probabilities.)I
fnd It extremely difficult to imagine a mass kili
butchering sita being avoicled by predator-
scavengers aner man left, so Ibal all bone
modifications at such a site were exdusively refer-
able to human achon. Yet Frisan has inlerpreled all
aUrition and breakage as a resull of human butcher-
ing, and has reconsum;led a form of butchering to
accommodate alllhe obvious attrition. Figures 3.3-4
and 3.35 illustrate a sheep pelvis, from a Navajo
winter site [L. R. Binford and J. B. Bertram 1977), that
has beeo gnawed by dogs. IlIuslratad in Figure 3.39
are carihou innominates Ihat have been gnawed by
wolves. It is certainly difficult to distinguish Ihese
bones from those cited by Frisan as evidence for
muscle-stripping butchering techniques. (Sea Frison
1970:17-18, Figures 11 and 12). Fimdly, the as-
semblage composition at Glenrock is consislent with
a kilI butchering site where anatomical elements
were being removed, ralher than a site where meal
was baing processed off the bones for transport (see
L. R. Binford 197Bb:475-476). Extensive muscle
stripping would only make sense if one was aUempt-
ing to reduce weight by removing meat from the
bones and in turn abandoning Ihe bones. As pointed
out, (he bones from the mosl meat-bearing parts were
nol abandoned al Glenrock! I am left with the con-
dusion that I agree with Frison that nmscle slripping
is one way of butchering that could weJl have been
practiced. and we must attempl lo find diagnoslic
criteria far its recognition. 1 musl remain skeplical
thal the facts from Glenrock and other siles supporl
Frison's interpretation.
In later reports, Frison acknowledges thal "r.arni-
vores can and do chew off musde attachments on a
cateass and it is necessary to be able to recognize Ihe
difference belwep,n ones Ihal were rflmoved as Ihe
result of bulchering process and those removed by
cernvores [Prson 1974:281." His reconstruction of
butchenng events et the Casper stte (Frisan 1974) is
much more conststent with my experencee, end
many of the types of breakage of which 1was skep.
cal al the Olenrock site are either not present al
Casper or they are not feetured in Frison's recen-
struction. Al! in all, it s my judgmenllhal great cara
must be exercised in using Frison's butche-tng
Breakage patlems coupled with anatomical seg-
mentalion and cut marks have served as Ihe basis fm
Frison's reconstruction. Less satisfaclorv dala have
servad as the basis for reconslrucling a-bulchering
procedure exdusively from breakage pslterns (see
Lyman 1978J. This interesling paper begins analysis
agansl a backdrop of previous research in which il
was largely postulated Ihal butchering had been ae
complished in Ihe southeastern Washinglon Stale
area almost excJusively by breaking bones. "Appar.
ently there is Iiule discernible evidence that bulcher-
ing the carcass induded a cutting process ILyman
1978:21." Guiding Lyman's analysis were the follow-
ing principies:
Brumley(1973, p. 241 believesthBtthe ways alldplaces
in whieh a bOM is brokell reflec Ihe melhods IInd pur-
posas of the brcakage. Gralltill8 Ihis the osteo
archlleologill! examines tha archaeologiclll bone as-
semblllge to define butcherin,R llnils. Butchering unils
are determined by eXBmination of bone units, braahge
pBtterns and bone groups. Abone unUis defined asany
fragment of bone. Brellkage patterns are defined as Ihe
eommon poinls of frllclure. Abone group i50 definedos
the Bssemblage of bone units representing a pllrticulor
skelelal elemenl, such a5a right femur or a left radius.
Boneullits, breakagepa\lerns, IInd bone groups arecon
sidered to be altribules of Dutchering [Lyman 1978:4J.
Lyman's procedure was lo describe each skeletal
element (e.g., scapula, femur, melacarpal) in tarms of
the "bone unls" into which il had been broken and
Ihen present the frequancy of each unil in differenl
archaeological populalions. No attention was given
ta cut or chop marks. wilh the implication Ihat none
were presento Afler this information was sum-
marized, an inlerpretatioIl was offered:
butcherin.ll consistedof Ihwugh Ihe limbsjust
proximal and dislal to Ihe joilllsof the long bones.The

suggest that the aboriginal peoplee of the southeast-
ern pert of Washinglon Stete were truly "culturally
unique." ,_---- "
I have observed the Nunamiut bludgeons'tc
break leg bones. particularly frozen baes, but such
a techmoue was only used on lower Iegs, where little
muscle sheathed the bone. More specficelly. hlud-
geons were used in breaking Ihe distal tibia just
aboye the articulator end, and the distal radio-
cubttue: less commonly they were used ust below
the proximal ends of the metapodials. Among the
Navajo 1beve observed ribs and transversa processes
ofthe lumbar vertebrae broken offduring butchering
wilh sn ax. Among Australian aborigines, 1have ob-
served a bludgeon-specifically, a Ihrowing stick-
used to break Ihe ribs loase from the vertebrae (Hg.
ure -4.43).In bolh cases ofbulchering with bludgeons
Alyowara usins o 'hrnwins S,if:k D bludSf'on in butcherlng.

' .
4. Human Modesof BoneModification Gnawing by Humons
Wolves, dogs, hyenas, and other nonhuman car
nivores do not have a monopoly on Ihe ael of gnaw
ng. Humans can, and most cerlainly do, gnaw
bones. Although bone gnawing by modem man is
relatively rare in the inlensive sense documenled
earlier for Ihe canids, we mighl expect some of our
earlier ancestors to have engaged in somewhat more
enlhusiaslic gnawing, particularly during the very
early time ranges when many different food
preparation techniQues were not yet developed or
used and parlicularly when our aneeslors were lak
ing Iheir firsl sleps toward regular hunling and meat
Are Ihpre ways of dislinguishing belween bones
means of accounting Ior missing borre elements by
the "fether" of bulchering analyss twhtte 1952,
1953a. 1953b, 1954, 1955). The conlinuaton of ths
nterpretetva tradition by Frisan (1970, 1971, 1974,
1978) and other Plans ercbeeologtsts [see lohnson
1977, 1978) fosters the view that chopping was e
major butchering slrategy.
As has been iIlustrated in Chapter 3, many of the
missing bones Ihet White explained away by a "de-
structon through butchertng wtth B chopper model'
were probably not preserved archaeologically be-
cause of consumptton by bcth wild and domesttc
canids. Similarly. mallY of the faels of destruction
cited by Frison and others as evidence for chopping
may well be gllawing destruclion by canids. As will
be shown. many of the patterns of bone breakage
cited by White (1953b. 1954. 19551and Frisan el al.
(1976j as well as Lyman (1978) as evidence for chop-
ping as a butchering tactic could easily be seen as Ihe
consequence of breaking bones for marrow, or, in the
case of Lyman, pounding bones for bone rease.
It is clear lhat chopping is documented to have
been used in butchering animaJs, but there is no
methodology adequate for recognizing the canse-
quences of this IIICUC, in either forms of breakage or
bone destruction, that is nol ambiguous ami subject
lo alternative inlerprelation. This is an arel' in need
of stroog melhodological development.
Gnawing by Humaos
technique of rib eemovel is lo cut Ihe distal ends of
the ribs from the brisket. seprate the main rib slab
Ircm !he rst twn rtbs, which normally remain al-
tached to the first two Ihoracic vertebrae. and !hen
grasp the rib slab elong its distal margin and pull up
smartly. breakng the rib heeds at their point of ar-
tculation wilh the vertebrae. [See L. R. Btnford
[1978b:48--60.94-971 for further butchering descrip-
tions.) With the moose and with large bear that 1
have observed. the fcregctng technique was alterad.
The first five ribs of Ihe slab (ribs 3-8) were broken
off at the proximal ends one al a lime, which re-
sultad in a very oblique break yietding a very
poinled broken proximal end ol the ribo Once the
smllller ribs were approached, Ihe butcher grasped
the distal margin ol the rib slab and broke it back all
at once, as is Ihe normal rnelhod with smllller ani-
mals. Once Ihe breaks were produced, the knife was
mn between Ihe breaks as the slab was pulled up and
il was removed as a unit.
The final accommodation of the Nunamiut bulch
ering technique to body size was in tbe segmentation
of the vertebrae. The '-nife was used ID cut ioto the
articulalion between the sacrum and Ihe lastlumbar
vertebra. Once sorne of Ihe conoective tillsue was se-
verad, Ihe entire spinal column with attached pelvis
WIIS placed uprighl. with Ihe posterior end of Ihe
pelvis resting 011 Ihe ground. The hunler Ihen leaned
down on Ihe vertebral column while pushing for-
ward, pladog a strain on the articulalion between
sacrum and lumbar vertebra. This normally resulted
in cracking Ihe articulation. permitting adhering
connective fissue to be cut readily with a small knife.
This same technique was used in segmenting the
calumn between the thoracic and lumbar verlebrae.
These and alher accommodations lo body size
rnakeit appear unlikely Ihal "chopping" is a stralegy
Ihat is necessarily related to large size, as is fre-
quently suggested. Conversely, as has been men
Honed, in my few ethnographic experiences wi!h Ihe
tedmique it was used on relatively smallllnimals-
sheep and bngaroo.
The assumplion that the use oEchopping tech
niques was widespread has been largely foslered by
lur.ctional inferences from lools thal subsequent re-
search, such as Ihal melllioned earlier by J. n. Clark
andC. V. Haynes (HJ70], has cerlaioiv l:Bst in doubr
FinalJy. Ihe f1%unlJltion of chopping W{lS (l post hoc
from the humeral shaft. As the proximal humerus-
scapule articulation was approached, two rether
quid. cuts were made "nsde" the stripped muscles
and around both the lateral and medial sides of the
ertlculation. Once these seemingly superficial cuts
were made, the butchers foot was placed on the dor-
sal surface of the scepule-humerus joint. The foreleg
was then given a rohust twtst away from and then
back toward the body, eastly disloceting the ont.
Once ths wes done. several minor cuts of connective
tssue freed the forelimb from the shoulder. The
scapula was then removed from !he thorex by cutling
between the lateral surfece oEthe rlbs and the medial
lJurface of the scapula. which freed it handily as is
the normal case when the entire fronl leg is removed
from smaller animals.
lt will be recalled that with carioou and sheep the
FIGURE 4.44. PurliuJly h u l c h ~ r f l d femule mUfJSfl.
or using a chopping technque, the stee of the animal
was reletveiy smatl (e.g., sheep end kengerooj.
Among the Nunamiut, when a larga animal is butch-
ered. slightly differenl tecbntquee are used, which
employ more leverage and more dsloceton of jctnts
coupled with cutting rether Iban chopping. For in-
stance. Figure 4.44 shows a emale rncose partially
butchered by a Nunamiut hunter and his stster. The
animal weighed approximately 780 lb end was com-
pletely butchered with 8 penknife. Sorne butchering
dterencee, which were obvtoue accommodetons to
the sze of the animal, relate lo the removal o the
front leg, Iha removal of tha ribs, and Ihe segmenting
of the vertebrae.
The "large-animal" stmlegy of front leg carnaval
WIIS to strip the muscles from the humeros by c:uttillg
insertiolls at the humerus-radio-cubitus oinl away

/w..) - ""'t.&r' C$lf "Vx
4. Humen Modes af Bone Modifkolion Mcrrow-Bone Breakage
gnawed by humane and thoee gnawed by animals?
As 1sud rny students contine to enelyze the control
data collected among the Nunamiut, perticuleely
faunel matertels recovered from houses end from
undoubted human use ccntexte. 1feel condent that
1 will be able lo provide al leas! sorne descriptiva
data reletve lo the pattems o borre moditcatlon sud
rleslrucllon generated by gnawing Eskimos. 1 have
mede sorne ethnographic obeervettons thet may ad
in Ibis work. For Instence. It was noted previously
(L. R. Binford 1978h 151-52) that gnawing and
"sucking" o short secttons of ribs was 11 fairIy como
mcn Be! where fresb meat W8S beng consumad, but
when dry andfor frozan meal WIIS beiog eaten steps
were generally taken lo preven! breakage o 5uch
parts as ribs, since aven Iba small amOllot of blood
and "marrow" in such bones is genarally not very
8ppetizing. The gnawing and sucking on rb "lItb-
lets" (sectiom of rb mnging from 2 to 6 inches in
length) is almost exclusively restricted to the ends of
Ihese tablets. Scoring, punctures, Ofpitting along the
lateral margins of ribs wefe never observad. This is
certainly different from what canids do to ribs where
they regularly crenulate the edges, as is sbown in
Figure 3.33. In addition to ribs, 1 have regularly ob-
served buman gnltwing on dorsal spines of tboracic
vertebrae, proximal margins of the scapula, snd var-
iDUS very "soft" bones of yonog individnals. In tbe
few iocidences where I have observed the conse-
Quences of this goawing io any detail, the pittiog
and tha "mashing back" of edges is fairly cornmon.
Punctures were rare aod scoring was never obsenred.
This ls an area where diagnostic properties might
well be developed through direct cJ(perimentation
with modem subjects. However, Ihe gnawing would
haYa to be dlreded and conlrolled for the slrength of
the subjects and they could not be allowed to selacl
what they wanted lo gnaw. Inslead, theexperimenter
would have to jnstrud his subjecls lo gnaw assigned
anatomical parts in specified ways so as lo oblan as
complete a picture as possible as lo wbat is possible
and Iikely under differenl conditions. 1 feel it is
highly unlikely that a normal homlnid pattern of
consumption would inelude gnawiog to Ihe exlenl
that iI would mimic the consequences of Ihe elassic
carnivore's behavior, as is documenled in Chapter
3. Neverlheless. we need dala on Ihis problem befare
tooth modtcetions on bones can be assigned lo
nonhumen egents in a totally relteble rnanner.
Marrow-Bone Breakage
As indicated in Chepter 3, pettems of bone breek-
age have served as focal ponts of concern for many
who have ettempted to recogmze human occupa-
tons when convnclng evidence in the Iorm of slone
tools was lackiog. I have thus far ergued that th!!
crack-and-twist method is nol essential to sp:tl:eUiac-
tull!. This optJf:fi'ili'
Similarly, the position Ihal SpifBI
fracture itself was a characteristic diagnostic of man
has beeo dispalled by my discussions of bone free
tUfe by animals and olher agents. Neverlheless, man
does characteristicaUy break bone for marrow, and
we would like to lmow if thera are properties thltl
would permil us lo recognize bones broken by man
as opposed to other agents, or bones broken by rnao
in the context of butchering as opposed to IDarrow
eJ(trBction or perhaps olher conlexls such as 1001
As mentioned earlier, a study of marrow-bone
breakaga by Zierhut (1967) with follow-up observa-
tions by Bonnichsen (1973) is to my knowleclge the
first specifie behavioral observatioo made 00 Ihis
subject. The observations by Zierhut were among Ihe
CaUing Lake Cree Indians of Canada. The observa
tions were made in Ihe "yard" of a permanent resi
dance where bones of moose had been introdueed
from hunting successas in the local area.
Tha meal is finl strpped from the hanes from whlch
marrow is to he blken. Onu Ihis is accomplished Ihe
bones are thrown on or next la Ihe coals of an openfile.
They are healad in Ihis manner for a very short time,
being lumad over occlIsionlllly lo prevent burning. The
Indians relale Ihat this procedure makes thll bones
aasiar to break. Likewlse.fresh bones break more
thltn do dry bonas. Nexl Ihe bones are taken {rom the
coaJs and allowed lo coal. Meanwhile Iwo large oval
stonas are placed on the groond about 8-10 i!lches
spart. Whan Ihe bone is cool enough lo handle, il is
placed so (hat Iha proximal and distal ends upon
the stones. the midshaft te thus suspended between the
srones. The bone is then struck in the middle. usng Ihe
blunt end uf 11 amali IDe. Wilh anly one or two vertical .
bowe Ihe bone s broken tnto two major beivas: durng
this operatlon a fewsrnall fragmentsor chips uf bone are
generally detached fram the poinl 01 impact [Zierhut
These ethnographtc experiences impressed Bcn-
ncbsen lo the poinl thet he seemngly accepted the
Cree behavior as being the cherectertsc mathod of
merrow-bone breakage. Bonnichsen (1976:36) relers
tothis as Ihe "rnid-diaphysis smesb technique of spi-
ral fracture." It should be pointed oul Ihat although
no olhar elhnographic descriptions were apparenlly
available, the actualistic sludies of Sadek-Kooros
(1972) c1early indicaled ways of breaking bones
other Ihan that observed amang Ihe Cree.
Sadek-Kooros (1972, 1975) conducted sorne fairly
extensive experiments in bone breaking; however,
the overall thrust of her research was a "replicative
experiment" designed loargue in favor of abone 1001
assemblage at 'aguar Cave, Idaho. She slates,
During IhE! identification of the Artiodllclylll. a consis-
leney in the shape and fracture paltems 01 many frag-
mants. particularly among Ihe long bQnes. was noted.
While il is nol inconceivable Ihal hane would tend lo
frscture alongUnesof least resishmce,that s. following
in!laren! slresses and slrans ... yet Iha possibility
existed that such pllllerning reflected purposerul frac-
luringaiml:!d st oblalningspacificshepes ISadekKooros
The parlicular pattern Ihat atlmcled Ihe laguar Cave
research was whal my field crew in AJaska referred
lo as "bayonet" breaks, described as fallows:
6 metatarsal bone relainiflg one epiphvsis. usually the
disfal fractured regularly the enl;re
posterior or antflriorsurface at specific dislam;:es fram
Ihe epiphysis retainad, Ilmi hsving a shaft at leasl J
limesas long as il was wide on Ihe surface apposad to
Ihesurfa<;;e fractured ISadek"Kooros 1972 ;:1711
This type of break was oblalned by placing the
metatarsal on an anvil (see Sadek-Koaros 1972: Fig-
ures 1, 2, 3) and hilling it, wilh
one or severel conlrollooblows(wilhcbtuse point of the
hammerstone] adrninlstered on Ihemedial andJorlateral
surfece ciose to the midpoint of the sbaft. followed by
Iwisting the distal eppbyels. The use of lesther or cloth
wrappad around the hammerstone or around the bone
being fracturad helpad reduce the shattering effect of
Ihe impect. control the force af the blow and regulate
the extent of the fracture [Sadek-Kocros 1972:3711
Although thi& hiihn'!flle was of primary nterest lo
Sadek-Kooroe since t permitt9d her to approximate
the forro of breakage she had charectensncetly seen
in her archaeological rernains. it is of sorne interest
thet she reccgnzed two cther Iypu. o in
addition to spiml fracture: green-slick fraclure, pro-
duced when lhe bone was slruck across a stone anvil
or another bone, and tronsverse fracture produced
when lhe bone was broken by bendins down bolh
ends across a slone anvil, another hone, the eJ(peri-
menler's knee, and so on.
Now, although Sadek.Kooros docs nol report the
delails regarding the fracture procedures that were
foUowed. it is clear that they were suecessful ways of
breaking abone, even Ihough they did not yield Ihe
forms in which SadekKooros was interested.
Could 1argue about Ihe way men of the past broke
bone by malching paHerning in an archaeological
site with the debris prodllced Sadek-Kooros using
"green-stick" and "transverse" bonebreaking pro-
cedures? Would Ihis be an argumenl from analogy?
If we are attempting to develop methods far distin-
guishing bone broken by humans from that broken
by animals. are we not interesled in underslanding
the range of variability that human bone-breaking
might lake? Is not Sadek-Kooros human? Is it nol
reasonable Ihat lhe tactics she usad to yleld palterns
of bona breakage differenl from a elassic spiral frac-
lure are also Iikely lo have been employed by other
humans at olher times? I think we must answer yes
lo al! these questions! Workers like Bonnichsen
would apparently answer no, since lhe only lech-
nique discussed ar considered as characteristic of
human fraclure by him was Ihal performed by the
Calling Lake Cree, the "mid-diaphysis smash tech-
niQue of spiraJ fraclure lBonnichsen 1979:361." Is the
only form of valid arBument from analogy elhno-
___bserved the Nunamiut Eskirno
break marrow J will place theee descriptions
in the sftuatfonal orsystemic context in which !hey
were most often observad.
1might unmrnent al ths point that in my previolls
descriptions I mentioned thet the Nunam
t" re
h 1 ( 'JlflslC ..__ gktxts (L. R. Binford
1978h:153), but Imada no systemettc attempt ether
lo survey rny notes sud extrae! the detals 01 these
differences or to mennon the total renge of bone-
breaking behevtor that I hed in fact wuneesed. 1was
very intent on treetng butchering in detetl bu! al the
time t did not occur to me la treet bone breaking
with equal attenton. 1was enemptng lo understend
interassemblage variablity in faunal remains con-
siderad from ao anatomical perspective. I was sim-
ply not asking the questions that 1 am now posing
regarding the relationships between the
ofbroken bone and the tlH::tcs used in bone breaking.
In many of the following descriptions 1 do not have
the bear and Ihe foolprint together; most of the time 1
observed Ihe bear-Ihe behavior of the Eskimo--but 1
did not think to record its footprint, the resulting
patlerns obone modification.ln short, much ofwhat
is to follow is ethnography, not ethnoarchaeology.
Marrow Cracking during Butchering Episodes
It is not uncommon for a butcher to crack leg
bones open while engaged in butchering. A como
ment that is frequently made is that you can teH the
nutritional state o the animal by examining Ihe bone
manow. In spring and doring 88l"ly summer, when
nutritional conditions of caribau are generalIy poor,
a bUlcher may break the front leg just aboye the distal
articulator end of the radio-cubitus. This is done be-
fore skinning and whHe lhe leg is fully articulated to
the animal. The procedure is to grasp the front leg,
bringing Ihe foot and metapodiiTtOrest on one's
Ic.nee. while standing behind the front leg fadng the
teaa of the animal. With the left hand the foot is
pulled upward and forward; the butcher is still bear-
ing down on the leg against the knea, In the right
hand Ihe butcher has a (frequentiy the
lower leg of anolher caribou). w Ich is used lo de-
liver a heavy blow down and on the poslerior edge of
4. Human Modes 01Hone Modificalion
the lateral side of the distal radius. One blow Is gen-
erally aIl Ihal s naeded. A skinning knlfe is then
used to cut tbrougb the skn and tendon al the break,
freeing the artculated lower leg (phalenges.
metacarpal, carpels. and distal radlo-cubitus]. which
is tossed to the side after the marrow s removed
from tbe distal end of the radfus. On the other hand.
if the break was loo high. resultlng in a eonairierable
length of diephyeis remaining on tbe distal end, the
butcher might use the handle nf his sknning knfe
and stnke at a protruding sectcn of dlapbyss. wifh
the intent of drivlng off a flake longitudinally down
along the bone. thereby exposing more of the mar-
row cavty.
Under normal conditions, the break remaining DO
Ihe diaphysis o Ihe radius and arm of the cubitus is
vel}' sharp and dagger-like, with a classic spiral
moving up and over !he posterior of the face of the
radius. The lower leg would be left allhe butchering
site and the upper part of the front quarter would be
transported to the residential location or to a cache
from which it would later be nlroduced to a residen
tial location. The results of breaking through the
front leg during bulchering are well recorded in the
inventories of anatomical parts remaining on kili
butchering locations (see L. R. Binfmd 1978b:76),
where without exception the discrepancy belween
the frequency of proximal and distal ends of abone
in the front leg is greatesl for Ihe radio-cubitus.
Less commonly, Ihis same behavior may be di-
rected toward Ihe lower third of Ihe distal tibia wilh
the intenl of removing the lower rear leg (phalanges.
melalarsal, tarsals, and distal tibiaj. At least among
the Nunamiut, cracking Ihe tibia is considered more
of an indulgence and is only done when the butcher
desires to eat good marrow. It is not a "testng" tech-
nique like breaking the radius. The only difference
in the way lhe tibia is treated is that I have observed
iI broken after skinning; in fact my notes indicate in
79% of Ihe cases 111 out of 14) Ihe rear leg had beeo
skinned prior to breaking the tibia for marrow. Oth-
erwise Ihe procedure ofbone fracture was essentially
the sama as for the radio-cubitus. On Ihe other hand.
removal of marrow was more extensive and more
atlention was given to modifying the shaft remaining
atlached lo Ihe animal. For nstance, in 3 oul of 14
cases, after the lower leg had been removed by cut-
Marrow-Bone Breckoge
tng Ihe tendons and tissue eround the break, the
butcher walked to a nearby rock and hit the protrud-
ing fractured end o the distal tibia on the rock. The
compect blob of marrow from inside the distal tibia
fell out neetly onlo the rock. Tha butcher picked it
up. putttng it into his mouth as he tossed the lower
leg lo (he sde. In 5 out of 14 cases the butcher par.
tiafly fUleted the mear from the tibia. leaving it at-
tached to the mass al unlleted meat en the femur.
This had the effect of expostng the main pert ol the
tibial shaft up te the pont al which {he shaft wdens.
just below the proximal end of the tibia. The butcher
Ihen screped the proximal surface of the tibial shaft
and slruck a blow directed at the proximal face of the
tibia about halfway between the previous break and
the proximal end. In all 5 cases a classic spiral frac-
ture deveJoped, running in both directions so that
several splin!ers could be pried up and the "hot dog"
shaped section of marrow Iifted from the bone as a
single unit. The fracture tended to disappear into the
uncleaned proximal end of the tibia still surrounded
by substantia1 quantities of tissue al the point of ar-
ticulation with the distal femur. In all but one case
Ihe bUlcher then picked up a tabular stone or a lower
le8 secHon to use as a bludgeon. The slump o the
broken tibia was then held and pounded with the
bludgeon on the fracture. Several subslantial blows
were delivered, the bludgeon was tossed to one side,
8nd the fracture was examined to determine if loase
splinters had been produced. The knile hlade was
inserted in the end of (he bone and used as a lever
until a section of bone gave way. Then Ihe blade was
run down the outside of lhe bone along the indicated
crack removing the periosteum a10n8 the crack. The
bone splinter was then bent outward, generally eav-
ing Ihe proximal end of Ihe splinter stiB attached to
the proximal end of the tibia by the periosteum and
alher lissue. The bent-away spllnter "door" then
permilted the easy removal of Ihe marrow in the
proximal end ol the tibia. The proximal articulalor
end wilh bent-away 5plinlers still adhering was left
with the heavy muscled "ham." which was eilher
later {ilIeted or carried away as a unit to a residential
If any further marrow was eaten during bulcher-
ing it was almost invariably done during a resl and
Ihe metatarsal WijS the target bone. Fracluring of the
metetersal under these condttons was eesentially
done in the sama way that il ts normally done in
hunting cemps and stands and will be described in
that conlext.
Marrow Cracking in Hunting Stands and
A3 I heve deomonstreted, consumption of bone
marrow s a common snacking strategy in hunting
campe and stands (L. R. Binford 1978b; 179-191).
The most consstent terget of this consumpton is the
metatarsal. 1observed 61 cases of breakage for mar-
row of the metatarsal in this contexto
The first point to be made is that bone-breaking
strategies in the field are variable. unlike Ihe situa-
tion in residential camps where a redundant set of
bone-breaking pattems was evident. The variability
in field sites is primarily conditioned by the contexl.
Mosl consumption in hunting camps and stands is
snacklng and is done quite expediantIy and as a par
tial function of the soclality of the moment, when
several people are snacking and talking logether.
Dnder such condidons, one looks about both fm a
source ol snacks as well as t001s for processing bone.
Tha occupants Dthese field camps, are usually men,
and the knives they regularly carry expedientlyserve
to break marrow bones; the blade is held and Ihe
bone is struck with lhe knife handla. Figure 4.45
shows the deaned marrow bones of a rear leg to-
gether wlth an anvil, a stone hammer, and the skin-
ning knife commonly carried by today's hunters.
when a skinning knife is not used for breaking mat-
row bones, an elongated stone such as the ene illus-
trated (Figure 4.45) is used. The anvil. although
sometimes used as such, more commonly serves as a
working surface on which marrow recovered from 8.
bone may be placed or accumulated prior lo con
sumption. It may also serve for tapping Ihe marrow
out of sorne articulalor ends. Typically the marrow
bone is c1eBnedto lhe stale i1Justrated in Figure 4.45
before it ls hroken. This will vary, however, with
whelher snacking is a leisure activity earried out 00-
Iween work episodes (such as butchering, hunting,
Of transporting fieldbutchered parts to storage or lo
Ihe residence) or <In activity done while a worker is


> .....
. ,
152 4. Human Modes o/ BOlleModifiClllion
MarrowBone Breckcge
FIGURE 4.45. Bones ond tools "sed in Nunamiul marrow crocldnR
FIGURE 4.46. Dcmonstrolion of impac' olJgllmelll for bOlle and hommer when a m{'lolonool ;5 being froclured.
series of bone platas. whieh can then be picked off
since Ihey normally edhere to the bone by virtue of
small segmente of ussue. 1use the term plates sfnce
the thckness of a caribou metatarsal is nol great and
the thickness ts uniform over mueh of the shaft. If Ihe
blows are deltvered with the flat side of a ham-
merstone or wilh a knlfe handla wrapped with a
smelltutt of skn. the blows are dtuse and have the
effects of "depressing" oul bone platea mueh Itke Ihe
segrnents o bone compostng a "depressed fracture"
of the skull or other large bone. Mosl of the time
there are few lf any impael notcbes or classic prop-
erttes of a percusson poinl mpact with tbis melhod
of cracking the metatarsal. The remains of the shaft
are elso quite dtstmctve. generally exhibiling verv
angular profile of fracture, as shown in Figure 4.48.
lf the metetersal is disartieulaled from the tersels. the
witha small suck or. if the knife handle s being used
asa hemmer. with the blade of the knife. Somelimes
the broken distal end may be tapped on tha anvil so
fhe marrow will pop out. Ths may result in "re-
touch" 00 the apex of the bayonel break.
The distal end wth ettached large sectton of
daphyss. or bone shaft, is Ihen normetly reversed in
the hand so Ihe proximal end o the bone te most
distonl from the body if it Is a left metetersal: a nght
matetersal would be postoned with Ihe distal end
away. The reason for ths is that Ihe lateral side of a
caribou rnetatarsal s much wider and slopes ob-
Jiquely up te the fluted crest on the anterior face of
the bone. II is elcng the flat faee at tts juncture with
the anterior chennel thet a series of quiek blows are
drected. Usually three hits about 2 inches aparl are
sufficient to crack Ihe "crest'' ofthe metetersal into a
right hand end a blow is struck clown on the aligned
"ridge" ebout 21/ - 3 luches back from the distal end
of the bone. Figure 4.46 shows [ohnny Rulland hold-
ing the stone hammer and metatarsal in the proper
posttlon for striking the bone. Figure 4.47 ebows the
spiral fracture resulting from the neck-directed blow:
the distal end is twsted off. since the peetosteum
held the bones together after the fractures from the
hammer impaet developed. TypiGally what happens
js Iha! the fracture runs down into Ihe artieulator end
on Ihe surfaee exposed to Ihe blow and runs away
and up Ihe bone on Ihe face opposite Ihe impael area.
This results in a poinled or shart "bayonet" projec
Iion of diaphysis along Olleside on the distal end of
the bone. The marrow cavily in the dislal end of the
melatarsai is smooth and Tlol webbed eaoeellous lis-
sue. Any marrow insirle !he distal end is picked out
engaged in cther tasks. Under the letter condtton
the bone may be broken while articulated lo others.
or it may be even broken while it is only partially
Under normal condttons of snacklng, when other
ecttvttes are not being conducted, the bcnee will be
cleened as shown in Figure 4.45 and placed up clase
lo tbe coels of the fire lo be warmed befare betng
crBcked. A metalarsal is Iypically grasped by a
righl.handed man in the left hand. while Ihe hammer
is held in Ihe right. The bOllewould be held slightly
lurned outward so Ihal if 'he melalarsal is from Ihe
left Ihe wide ridge 8100gIhe junc:tion ofthe posterior
and medial faces of Ihe bOllewould be uppermost. lf
a right melatarsal is being cracked the posterior and
lateral faces of Ihe bOlle would be uppermosl and
aligned to receive!l blow. The hammer is held in lhe
fiGURE 4.47. TwJSfillB off spiroJly frocturerJ dislal end of a merafoFSaJ.
FIGURE: 4.4e. Impact and fracture patlerns typicul uf mefapodiuls broken for JmI1TOW.
Le" metocorpal,
m'dial tcce

." 'W'--' . m""

'."- : . aHer ,mpoct re - /,.. Postll"or f<lCII,
'. \ - - zrio

RiQht metatarsal,
aleral lace

\, - .--
't: _---.co

,,,..,, tece

, __--- ---; -.....A 8
-- , - .;_ I =
\..,,-, - ,- ,..
lett metatars<l1
" Human Modes of Bone Mndi!klllioll
iIlustrated in Figure 4.46. However, tbe blow ts gen-
erally delivered with 11more rounded portien of tbe
hammer and with more force, so a point of impaet
about three-querters of an ineh in diameter is pro-
duced. The blow is dtrected somewhat lower on tbe
sde of the bone. Generally, before the distal end is
twsted off a seeond blow ts directed along the same
side of the bone about z inches in from the distal end
of the bone. Both of these blows develop clesstc Im-
pact scars ringed wth small. incompletely fracturad
ny impaet chips (see Figure 4.48). When twsted
eer the second hlow, the bone may well fall into
beaically four preces: a splintered sechon of the ano
tenor crest of Ihe diaphysis, the distal end.fhe prox-
imal end, and a sertlon of the posterior face of the
daphyse. This is a hard bone, so that most of the
dagnosttc properties of poin! impact such as dts-
proximal end may be split with segmenta of the
proximal articulator surlace edhertng to vertous
bone platee and spltntees. Once the fracturad bone
eplnters and pletes are picked off the bone. the mar-
row ts removed. normally by dggtng tt out with a
stick or with Iba knife blBde as ts shown in Figure
Another bone sometimes broken for marrow in
bUDtiog cernps and stands ts Ihe metaceroal. In
GBribou thts s structurally a very different Iype o
bone In that the dephyss is much thcker then in
Iha metetersel, and the rnerrow cavily s much
smaller. The bona is shorter than the meteterset and
it Is much more symmetrical in cross sectlon. Itjs
prepered in Ihe same manner as the metetarsel but it
i8 broken somewhat differently. The posttton of the
bone relativa 10the hand-held hammer is the same as
IJ"uoP. - --48<'" A... 4:"- j-.I
~ . _--'.
ttnct tmpact notchee remen 00 the fragments ot the
diaphysis. As in Ihe case of the metatarsal, Ihe prox-
imal end may sometimes splil out with Ihe sections
nf diaphysis.
Variability arises in the degree oC cleaning and
Ihe degrea lo which bones are disarticulated from
adjoining bones. It Is nol UDcommon to lee the lower
reer les of 8 caribou- phlcettup clOl'Ie lo Ihe-tire fur
warming pl'ior lo skinnlng, sometimea witt\ the -fool
snd larsal stiU articulated. Once Ihe bone Is warmed,
Ihe skin oC Ihe central par! o Ihe anterior BurJace
may be removed sud a series ofblows directed o.loog
Ihe uoctnre o Ihe latefal lace sud the anterior nute,
resulting in Ihe plates and thin bone splinlers de-
scribed for the other melhod of cracking bones.
These will be pulled up and commonly allowed lo
rernain aHached by lIssue at the proximal or dislal
~ "''1/'
....' ~ ..J
.,," "
4. Humen Modes of Done Modification
FIGURE 4.49. Removing marrow from me proximal secliQn 01 a broken melatarsol.
ende of the bone. A window ts thereby exposed and
warm marrow may be dug out. The result is 6
rnetatarsal wilh a Iypically angular fracture profle
along Ihe central part ofthe diaphysis bul with prox-
imal and distal ends remaining arliculated with the
fool and larsals respectively. Somewhat more com-
monly Ihe {oot will be disllrticulated and the bundle
of tendon Ihal runs aloog Ihe posterior surface of!he
metalarsal will be pulled up lo its altachment al Ihe
tarsals. The exposed dislal end of Ihe metatarsal will
be Irealed in the same manner as a complelely
deaned bone. The only difference is that Ihe prox-
imal end remains articulBted with Ihe tarsals and
assodaled with the fool bones, which are atlached al
Ihis time lo Ihe lendon bundle (see Figure 4.03).
The men explained thal Ihey used Ihe harnmer-
on-bone lechnique in Ihe field because it Is easily
Marrow-Bone Breakage
done standing up, it ls easily done so thal articulator
ende and the desred maITOW do nol fall to the
ground, and Ihey do not have to clean the bone per-
fectly or heve it totaUy disarticulated. The last Iwo
characteristics are importanl -elettve to the techo
ntques normally used by women (or processing mar-
row bones in residential cemp suatrons.
Men in the fieId may sometlrnes crack bcnes other
then the metepodtals. When thts is done they nor-
mally mpect the bones on the sidee. ettber lateral or
medial, with hemmers as was done Inr the metepo-
dels (an exceptton s the proximal fmur, which 1
will describe Ieter when I dtscuss strategtes em-
ployed in the reaidential camps). They always im-
pact the bones on the neck area jusi ebove the ar-
tculator end. With bones such as the distal femur
end tha distal tibia. they may impact the bone up lo 4
nr 5 inches from the erttculetor surfece. aimlng for a
long bayonet break. which exposes the medullary
cavty but does not crack the end off the bone. The
letter musl be twteted apart given the aUachments of
the peosteum. This ensures that the ertculetor end
does not fall to the ground or otberwse gel "out of
hend." Bones with dense daphyses such as the libia
!end to break with cteen "percussion" relatad prop-
erties, yielding splinters with clean lmpact notches
(Figure 4.50).
When meat is being cut from (he bones in the
eld for purposes of eher field drying or weight
reductton for transporl, complete legs or melor ar-
uculeted segmenn may be Iying around. with the
meet removed. Nol uncommonly a butcher oc persnn
engaged in filleling the rneet may greb one of Ihese
arliculaled but filleled legs and praceed lo crack
bones-generally those of Ihe uppec leg such as the
humerus, ar Ihe libia and femur-by slriking Ihem
obliquely in cenler shaH while Ihe bone resls on an
anvil. The bone is (hen twisled aparl and (he marrow
is dug out from Ihe remaining Iwo ends. This cesults
in frequenl arliculalions of dislallibia and proximal
remur oc proximal melalarsal wilh tarsals and distal
libia, cleariy indicaling Ihal Ihe bones were nol dis-
articulaled prior lo hMing been broken in miclshaft
foc marrow. As stated flarlier, bones impacted in
miclshaft lend lo show df:!veloped impacl nolehes
and classic spiral fractures. when Ihis slrategy is
used. the imlwct is nearly o/W<lYs 011 Ihe laleral oc
medial sides u( Ihf' bone, sin(;e it is diffic.ult lo
fiGURE 4.50. Pernnsion impoct notch on lonll-bone
streghten out the joints and placa abone on an anvl
wllh anterior or posterior facas uppermost.
Marrow Cracking during MeaJs in Residenc!
Sites ---
Most of the time marrow bones are filleted by Ihe
women while they are preparing rneals. and the
bones are atorad for later processtng in large groups.
On sorne occestons. bcwever, rnarrow bones may be
served at a meal (see L. R. Binford 1978b:145. 146).
When Ihis is done Ihey are almost always eaten by
men or al least broken by men, although women may
receive choice morsels from their male relatives.
Somelimes an oid woman will break her own mar-
row bonesaround a household hearth, and when Ihis
is done (1 is almos! always dona in "male style," Ihat
Is, using a harnmer lo impact the bone held in Ihe
hai;ld. It was explained that Ihis ensured Ihat the
ends did nol fall lo Ihe ground or in the hearth, since
Ihis is generaJly done in fronl of diners inside a
winler house. An alternative melhod is sometimes
IIsed nside hOllses when Ihe marrow bones have
been warmed in a slew and are very warm and slip-
pery. This is Ihe "on-anvil" technique, where the
bone resls on an anvil----cornmonly one of Ihe large
slones surrounding the hearlh-and is impacted
with a hammer. Hones broken in thls manner are
seeted 00 the anvl in a stable fashion and do no! slip
arnund. Bones commonly seated 00 the posterior or
anterior faces are distal metapcdtals. distal radio-
cubitus. and the distal tibia. ThsS8 are almos! always
impacted 00 the posterior or anterior faces rather
than 00 the lateral or medial faces. as is more com-
mon when the bone is held in the hand end impactad
with a hammer al en oblique engle lo the bone. The
latter ortenteton biases the break in favor of B cleestc
spiral fracture. which ir done well will run up the
cevty exposing a considerable portien. Impactlng
on the posterior or anterior faces tends lo shatter the
bone or produce a complex depressed fracture (com-
mon with the radio-cubitus and metapodials) or a
complex bipolar fracture more Iikely wlth the dstel
tibia. Removal o the artioulator ends in this fashion
commonly resulta in an unbrcken section of the shaft
of the bone in the forro of a cylinder (particularly
with the femur and tibia), whch te then Creedof its
marrow by pushing the "hot dog" out wth a small
stick or sometimes a sheep rib bone. Many women
are more comfortabla with the cn-anvtl technique
and 1have seen it employad by them at large migra-
tion hunliog slands (sea my descriptions oC behavior
at Anavik aod Anaktiqtauk (L. R. Binford 1978b:
171-1781l. Regardless of whether the hammer-on-
bone tel::hnique ar Ihe on-anvil technique is used,
the impal::t8are slill direcled al the neck ares of the
bone jUst below Ihe arliculalor ends.
Processing of Marrow Bones by Women in
ResidentioJ Camps
Most marrow-bone breaking in residential eamps
is done by women during intense processing ses-
sions (SBe Figure 4.51 l. Commonly Ihe marrow bones
ara col1el::ted for several days and then processed al!
at once. Occasionally when large quantities of bones
ate available, as during migralion hunting aclivities.
massive quantities may be processed alance.
Typically a skin is placed on the Aoor or ground,
depending where Ihe processing occurs (normally
outsidel. Do Ihe skin is placed a conlainer for the
marrow and en anvil slone, and Ihe bones lo be
crack.ed are pUed nearby. The woman is sealed al Ihe
4. Human Modas 01Bone Modi/icalion
FIGURE4.51. Mass uf bones oWlJllins crack;ns by Anak-
tuvuk women durlng "gong" processlns for JnalTOw.
edge of the skin, which serves as a drop cloth. Mosl
of the time Ihe bones have been previously cleaned
bul sorne minor c1eaning may take place prior lo
cracking. Actual breaking of the bones is done by
striking rhe bone down on a handheld "hammer."
The hammer tool is called a kootoh- and has beeo
illuslraled previously (see L. R. Binford 1979J. It s
typicallya natural tabular secHon of local Hmeslone.
relatively squate or Ihicker on one side than the
olher. This means thal by a shift in orientatioo, Ihe
kootoh can impact a bone as a point, a very thin
ridge. or a very broad, fiat face. It may also be shifted
from Ihe lefl lo Ihe right hand (if Ihe woman is
righl-handed) and used as a hammer io Ihe male
fashion if needed. When womeo break marrow bones
in residential locations, 'hey are concemed wilh
breaking Ihe bone so Ihal Iillle compact and dense
diaphysis remains atlached 10 Ihe articulalor ends,
which willlaler be used in Ihe making of bODe grease
on bone juice. There is a secano cOllcern: namely,
Murrow-Bone Brecknge
thet the bones be broken so thet the marrow remans
"oleen," thet s. free of impact chips commonly
driven lnto the marrow when the dense bone of the
stuLft s mpecred. Both of these concerns are served
bystriking sorne erttculetor ends as close as possble
lo the polnt al which cancellous ttssue merks the
transition from the medullary eavity lo the ar-
culetor end. Such cancellous tssue is typically
found in both ends of tbe femur, Ihe proximal tibia,
and the proximal humeros. On the otber hand, rele-
tvely srnooth tener surfaces are characteristc of the
distal humerus, distal libia, proximal and distal
redtus. and proximal and distal metapodial. In
carbou, the metatersal end the distal radius are rele-
tvely thn-walled and may be broken by impact
crushng the erea jusi aboye the distal artculator
end. This results in few mpect chips and only de-
pressed platee, which can be picked off the marrnw.
Only the distal tibia, proximal radtue. distal
humerus, and metacerpal are charactertzed by
smooth inner surfaces of the medullarv cavttv. and
thick and dense adacent diaphyses. bones are
ofless utility as grease bones beeause of their densily.
The degree to which Ihey will be broken close lo Ihe
end is Iherefore related to Ihe quanlily of bones
available for rendering grease. For instance, if many
bones are available, as in the case o Figure 4.51,
lhese bones will be struck obliquely in the secUon of
shaft judged lo be mosl perfectly round, since ob-
liquely impacting Ihese with eilher Ihe hammer-on-
buneor boneoflhammer technique, as ilIustraled in
Figure 4.46, tends to produce c1ean spiral fractures
running bolh directions wilh Iittle impact notching
and hence Ihe driving of chips inlo the marrow. Mar-
rowconlained in Ihe longer unborken secHon of the
dislal tibia, for instance, may be dug out; oceasion
ally the spirally fraclured end may be lapped on an
anvil. knocking Ihe marrow free. If this is the
strategy employed. Ihe dense bones, dislal tibia. dis-
lal humeros. proximal radius, and ends of Ihe
metacarpal would nol be pracessed for grease wilh-
out further reduction of the diaphysis. The lalter
would be done as part of processing articulalor ends
for grease rendering, ralher Ihan as part o( Ihe
marrow-processing aclivities. The technique of
bone-on-hammer cracking very clase to Ihe ar-
Iiculator end leaves very dislindive Iypes of breaks
as well as impacl areas.
Figure 4.52 tllustrates an Eskimo woman about to
strike the proximal end o the fmur on the hand-held
harnmer-envtl. Ftgure 4.53 shows the type of break
that resulte from a blow of this knd. The proximal
end s broken off, with HUle attached diaphysis.
Hints of slght depreesed fracturing of the thin-
walled bone are still vsble at the upper end of the
break. Impact Is nveriebly on the enteror face of the
proximal femur, just between the femoral head and
the greater trochenter. The fracturas are generally
oleen and have sorne of the properties of heving been
snapped off rather than sprally fracturad.
FIGURE 4.52. Proximal femur about tu be hit ocras," a
hond-heJd anviJ.

FIGURE 4.55. Pushing marrow rom a femaral cylinder
with a marrow pusher.
ken. In the letter case the femoral shaft would essen-
tially break open in the hand. If this does not hap-
pen. marrow from the cylinder is pushed out (see
Figure 4.55) end the marrow from the cavity of the
distal end s dug out both wilh a stck or somettmes a
more substantial rnerrow pusher, such as a lower
sheep ribo
The tibia is normally held by the distal end and
struck across the hand-held hamrner-anvil so the
impact is received 00 the posterior fece just below
Marrow-Bone Breokage
The distal end is impactad as shown in Figure
4.54;that ts, obliquely across the band-held anvil jusi
behind Ihe distal articulalor end. Impact is normally
on the lateral surfece. The curvature s relatively
sllght on ths surface and a massive impact zone s
developed with sorne of the properties of a depressed
fracture, with numerous small chips erranged in ra-
dial "tters" around the znne of impact (see Figure
4.53). Ashort spiral fracture normally comes up from
Ihe impact zone around the shaft on tbe side away
from the persono The person cracking the bone s
then left holding the cylindrical segment of the
femoral sheft if the spiral fracture does not intercept
B fracture produced when the proximal end was bro-
FIGURE4.54. Distal femur alignfld ayer hand-held oswtl
jU8/ befare impacto

1 1


i g i:
-- ,
-. "
E i
i5 .!








__ ::L..
4-\)o'X-....l. -
4. Human Modesof BoneModijkarln
Morrow-Bone Breakoe


the condyles. As is the case with the femur, ths bolle
Is relatively thin in this area and depressed fractures
remain along the margins of the actual fracture. Typ-
tcelly 811 the bone on tbe posterior face is broken
away, leaving cnly a pointed projection of the Iibial
crest. The bolle s then reversed in the hend end the
distal end o tha bone s etruck clown ecross the
hammer-anvil epproximately 2 loches hade from the
distal end o the bone.Hthe impact s done well, the
distal end s snapped off and falls lo the floor below
the envti. Classc spiral fractures are not so clearly
represented with ths method. If the bone s particu-
larly rohust it may be impactad across Ihe corner o a
stationary anvil resting on Ihe ground. In aoy event,
more Of less straight across but jagged breaks are
nonnalIy the resulto The tibia is senerally more ir
regular in section than the femur, so it is not un-
common to impact an unfractured cylinder using tbe
koofoh as a hammer, splitting lhe cylinder and
thereby exposing Ihe marrow.
The humerus is a shorter bone than either Ihe
tibia or the femur, yet the distal end is very hard and
dense. For this tellson the distal snd is normal1y bro-
ken off firsl. The strategy used is simlar to that used
for the distal tibia, except that the dital end is almost
always seated on a stationary anvil, medial face
down, and impacted with the kaotoh used as a
hammer Oh the lateral face directly on Ihe ridge
marking the anterior from the lateral face af Ihe bone
(see Figure 4.53). A strong bipoiar pattern of scarring
is cornmonly visible on the broken end. The bone
treated in this manner does not normally have a very
distinctive spiral fracture. Instead, a jagged trans-
verse fracture is common.
The bone is teversed in the band and the prox-
imal end is then impacted obJiquely down on a
handheld hammer-anvil so Ihe blow is lo the me-
dial face of the bone just below the lip of the large
proximal articulator condyle. Sometimes (his is
cracked off, using Ihe same strategy as for the distal
end, sinee (he butcher already has the hammer in the
righl hand and the anvil is in use. When this is the
case, the bone is sealed on the anvil wilh the edge
resting acrass Ihe laleral foce and the impact is de-
liverad lo Ihe medial face just below Ihe artculator
Fracturins of the rado-cubnus.ts more drectly
related to the degree o prevtous cleentng of lissue
from the bone then holds true for eny other marrow
bone. 1 have already mentioned thet it may befre-
quently broken durtng butchenng. This bone is
rarely served in restdental stes. since it is consd-
ered lo break nside the bouse.
Ifthe bone is cleaned the lechnique ts to strke Ihe
distal end of the bone down on the hand-held anvil
oriented obliquely, as was shown for the metapo-
dials (see Figure 4.46). Impaet is generally slightly
on one side, although it msy sometimes be on the
anterior face back about 1!h inches from the dislal
end. The distal end usually breaks off rather trans-
versely when this strategy is used, although the
break may be somewhat jagged. The bone will then
be reversed in Ibe hand and impacted 0(1 the hand
held hammer-anvil aHalong Ihe anterior face ofthe
radius, beginning very clase to the articulator sur-
face, and then hit at aboul l-inch intervals down the
shaft of the radius. Blows are generally jusi slightly
off center bul slill tend to be on Ihe anterior face of
the bone. This mode of fracturing has the effect of
breaking Ihe proximal end of the radius off very
short. Wilh matUte animals Ihe posterior face of Ihe
ladius remains altached to the cubitus, snd impact
searring may be overlapping down short sections of
the margins of the fractures on the body of Ihe bone
Marrow is primarily conlained in the medullary
cavity of the body or shaft of long bones. This shaft is
shaped like a eylinder, so access to the medullary
cavity and hence the marrow is facililated by col
lapsing or fracturing the cylinder longitudinally.
'PWLLSiBi 1I hus in lb n'erofJ'lns b PO b te
tnyrtm'piOfde lOOlt:v aeeass "" M16 gil"''',
If one simply breaks a long bone through the
center shaft, Ihen one must dig out tha marrow,
which is very Iime-consuming and not very satisfac
tory. or proceed to break the bone agaio, 10ngitudi-
nally, to expose the contained marrow. This results
in curved fractures on short splinlers. The most
likely bone form discarded is an articulator and with
very Htlle attached diaphysis. This is not fhe form of
bone breakage that is generally cited as deriving
from marrow-bone breakage and Ihe crack-and-twist
method. Breaking bones far lIH1rroW results in cer-
tain very regular patterns as observed among the
Nunamiut Eskimo. That what the Nunamiul do is
indeed a general strategv has been confirmed for me
bythe examinahon of many fauna! assemblages from
many dfferent time perlods and piaces where tnten-
uonel breakage by man is not in questton. The break-
ege strategy generally mvolves percussion impacto
whether it be (o) with a hammer on a hand-held
bone, (b) with Ihe borre nn a hand-held orrigid anvtl.
or (e) wilh a hammer while the bone re resting on an
anvl. Regardless of the loading tacttcs used. the
Iarget Impact zone on abone ts generally the same.
rhe marrow bone to be cracked is generally h"ld by
lhe most robusl articulator end and initial impact is
just below the neck of the moce compect articule.tor
end. This initial heavy blow normal1y results in con-
siderable fractures developed through the shaft of
the bone, and the resul1ing splinters are peeled baek.,
as they mey be held in place by the periosteum. This
peeling back exposes the cylinder of bone marrow,
which, if the animal is in good nutritionel condifion,
can be picked up and either ealen or added fo an
aceumulating pije of marrow. If the inHial impact
did not produce fractures extending the length ofthe
bone shaft, secondary impacts are mede along the
basa of the neck o( the opposite end. The force of
these impaets are generally less than the initial im
pael fhat froctured off the opposite articulator end.
These more modest impacls fracture the bone so
may once again be peeled or struck off lhe
cylindrical shaft. This strategy of direcling percus-
sion impacts to Ihe neek of the long bone (Ihe area
direclly below the articulator end) results in many
long-bone splinters and isoleted artieulator ends
with little attached diaphysis.
Control Collections
As mentioned earlier, these descriptions are cul
loofrom behavioral observations made while Eskimo
were breaking marrow bones, and 1 did not make a
practice of gathering lhe bones 1 observed being
cracked. Part of Ihis failure on my part was from nol
knowing at Ihe time what patterning had been ob-
served in the archaeologk,al record lo which my ob
sefvations mighl be of relevance. For nslance, at the
lime of my fieldwork, 1 was unaware of Frison's ar-
gument about bcne choppers. 11 therefore nevar oc-
curred to me to tnspect in detail the bones 1 saw
crecked for merrow during butchering when the
butchers hit the apex of spiral fractures on a rock tu
dislodge the marruw. In Iect. the potentlal impor-
lance of this act dtd not really regis1er unt I beceme
awere of Frison's observations. J took it for grented.
stnce I can remember discusstng marrow cracking
with Kent Flannery and he reoorted that this was a
common act in the Near East.
In order lo shakethe marrowout of Iheshafl of the bon",
the villifgers seemlo havetapped !he brokenend againsl
a nltrd surface. The brollan l:Idges of sorne limb bones
shaw a resultanl chipplng whch (allhough irregular)
almosl always rasemblesper(;usllion on flinl.
Dr. Junius Birdof IheAmericanMuseumof NaturalHis
tory reporls Ihal he observed Ihe Tndians of Patagonia
using the technique of marrowremoval when he visited
Ihem in the 1930's. Al thal time he noted the scaleJike
"'retollen" produced on Ihe broken edges of Ihe bone
This "retoueh" al50 oecurs on bones fran! AH
Kosh (see Fig. 123c, ellHole el al. 1969:2911
This illustrates that when one is studying living
peoples one simply does not know what is relevant
wilhout a detaHed knowledge of the paflerning that
has been recogniZ6d in the archaeological record
and, even more important, patterning for which in-
terpretative c1aims have been made_
1did not ignore marrow cracking, however, since
1 do have a number of colleclions o( bone known to
have been cracked for marrow and about which 1
have considerable informanl information, although 1
did not witness !he bone breaking. From a series of
such informant-documented assemblages, sorne of
lhe properties of bone fragmen(s generated by the
tscties described can be demonstrated. SCBes
froro hiltin the bone durin marrow crae og are
guite istmctive. Firsl, they are a mosl a ways at a
single impact point, which results in driving off
short bul rapidly expanding flakes inside the bone
cylinder. At the point of ilnpad the bone may be
nolched, in that a crescent-shaped notch is produced
in Ihe fraclure edge of the bone. (See Figure '1.50.)
These have the appearance of Claetonian notches in
the Bordean laxonomy of lithic fracture but are
somewhal smaller given the nature of the
bone. I h<lve never observad Iwo such impacts di-

4. Human Modes of Bone Modjfjrmon

TABLE 4.05
Bone Splinlers Recovered rom SiK MQTrow-Crockins
of Ihe bone; the chips rernain atlached because Ihe
fraclures are nol carded Ihrough Ihe bone. This is
well ilIustraled in Figure 4.53. (This patlern was firsl
described by H. Martin 11910).) Table 4.06 sum-
marizes Ihe frequendes of bolh irnpacl nolches and
depressed margins observed in Ihree popuIalions of
rnarrow-cracked bones collecled arnong Ihe NUDo.-
No. % No. % splinters
---- ----._---
Sample 1 59 .86 10 14 69
Sample 2 B3 .B' 12 .1B 75
Sample 3 50 85 9 .15 59
Sample 4 114 .84 za .1B 136
Sample 5 14. .63 30 .17 17.
Sample 6 312 63 B' .17 37B
Totals 742 63 147 .17 889

" o

" E
11 .e
Z "


".i::i.b ..

.... a>a> ....
.8 ; .......

11 iii iii
o o; al al t




" NOO<l'<O



"' ..
N ... ON'"
N_ ... O ...
N O""'" <1'
......... ",
.... OON'"
ii 2
041 u"g
o E Ol .f!! ......

..;i--- _E.c
.c: ._ ._ ..
.... aaao..

_ N
o I ........ N
"'NO .... ce
"'ON ... <I;>
"' ... 0"''''
N 100N
e .... ce r_ 1-
N_N <1'
.... O"''''' N
.... ONO'"
g -;1
- .
-g <i Ol

-oO __ Ol


0 ... 00_

N ... ", ... <O


"'''''0 ... ",


- -
N"'''' .... Q'l
e.. ;

t: o;; -;;-g

" <

, 1

I z






Sample desr-rfpttons
Sample 1 From the Mask site [sea L. R. Binford
1978b, 178-tAl); representa the ac-
uvutes of severa] hunters durlng a 2-
day perlod: blased in favor uf meta-
Sarnple 2 From marrow-cracktng demonslrations
Ireported in L. R. Binford 1978b: 155)
Sarnple 3 From the Kongumuvuk falJ hunling
stand (L. R. 8inford 1978h: 357);rep.
reseets the activities of lwo hunters
during one evenng
Sarnple 4 Bones seved Iur study by the women of
an E&ldmo household during ApriJ
1971; fre&hly Idlled animals intro-
dUcRnduring Ihe lasl two days oflbe
colleclion; hiased in favor o melapo
diaJs; relalive lo samples 5 nnd 6
Samples 5 and 6 Bones saved by Ihe head of an Eskjmo
household over a considerable peri-
od of lime during lale fall and earJy
winler; represent household con-
sumplion of marrow hones largely
froro stored meal; biased in favor of
upper limh parls (femur. tibia, hu-
rectly sirle by slde. This makes a greet deal oC sense.
stnce the nntches are only produced when the load-
ing successfully breaks the bone. If that happens
there is no further need for fracturing al that exect
spot. Any addilional hilling 01the bone will be along
an unfraclured surface. The frequency of such im-
pael nctches (Figure 4.50) in a marrow-creekad as-
semblage Is indicated in Table 4.05, which presents
the ebaervad frequencies of splinters [diephyss
fragments exceeding 2.9 cm in Iength) with Impact
notcbes in populations oC bcne splinters from six
control sampJes collecled al Anaktuvuk.
lt should be cleer Ihal the production of percus-
ston impacl scars on the diaphysis is a regular phe-
nomenon and can be expected lo exhlbit dstnctve
numerical pettemng depending on the mpact tac-
tos usad and the size and morphology of tbe bones
beng broken. For netence. in samples 1. 3, and 4,
where there is a bias in favor o metepodels. there ls
a slight reduction in the percentege of bone spl inlers
exhjbitng impact notching. Ths is conststent wilh
the earlier observations that fracturing, particularly
of the metatersal, was more like producng a series of
depressed fractures from yieldng bone plates Ieck-
ing impacl notches. On the other hand, essembleges
known lo heve contened more upper Iimb bones
such as the femur, tibia, and humerus clearly exhiblt
a hgher frequency of mpect-notched bone splinters.
Comparative study of Ihe frequendes of impacl
nolches is believed to be a fruitful approach lo dis.
linguishing between fraclured bone derived from
animal versus human aclioos.
Obviously the impacl scars are nol exclusive lo
bone splinters. They regularly occur on arliculalor
ends and io distinctive places, since Ihe placemenl
of blows is a regular phenomenon varying with Ihe
morphology of bones and only minimally wilh Ihe
loading slralegies. Impact nolches as ilIustraled in
Figure 4,50 are Ihe scars remaioing after a section of
bone has been removed by virtue of impacto There is
another type of impact scar that commonly occurs
when impacl is over a wide surface, when impacl is
on a surface "backed" by caocellous tissue, or when
Ihe bone IS exceedingly strong or dense and thick.
This is whal I have lermed a spJintered and de-
pressed margin (see Figure 4.53), Morphologically
speaking. this is cornrnonJy seen aloog a fraclure
where Ihere ma)' be sorne srnall segments of de--
pressed surface surrounded by a kind of roselle of
very small "chi ps" Ihal have (ractured on Ihe surface
.::;-wA t...'d -.e 1Z-dI

miut Eskimo. I have arrenged the arttculetor ends
into three baste groups ranging From those with
very canceltous Intertcrs through Ihose with smooth
but thin-walled m8ITOW cevtttes to those with thtck
and dense bul smooth interior marrow cavities....
111 ' L I tila: diG 3M:::':: ::616 a uf
o.b-. rr1._ 1 "f' tb t I h.. bLla
M.' t kh.'.
Several things are of tnterest in this tableoReview-
ing the lotals il is cleer that moving clown the teble
from cancellous 10 smooth end thick ntenors there
ls a correleted ncraase in the percentege ot ar-
ticulator ends exhibiting dislincl and oleen impact
notches. Only 8% were observed among the ends
udged cancellous, whereas 1B% of those udged lo
have thkk and smoolh marrow cavities 9l1ihibiled
impact nolches. Conversely. only 9% of Ihe Ihid:. and
smoolh ends exhibited marginal depressions from
points of impacl, wheroos 37% of Ihe cancellous
snds e",hibited such scars. There are several implica-
tions of Ihese dala. Clearly Ihe more cancellous lis-
sue Ihere is forming an "undersurface" lo Ihe bone.
Ihe more evidence of direcl impact there is remain-
iog on Ihe bone. Similarly.lhe more dense and com-
pact the bone, Ihe deaner the fracture and Ihe leros
direcf evidence of impacllhere is. Perhaps Ihis iB nol
as indicltlive a correlation as it might appear, since
good examples of marginal depressions and "chip
roselles" are in evidence on sorne long-bone seclions
froro biso" bone iJIustrated by Frisan (1974: 301.
Figure 1.13b, 1.24; 1976: 307. Figure 8.213). Similarly.
Ihe bovid bone impacled by Bonnichsen (1979: PIale
IV-2, No, 10) experimentally wilh the "mid-diaphysis
smash lechnique" c1early shows Ihe chip roseltes,
and the impacl depression characlerislic of my more
cancellous bones. This arrect may be more a funclion
of Ihickness Ihan of bone densily. In any event. the
evidence from impacling bone is common on artic-
ulalor ends, accurring on over 35% of lhe lolal num-
ber examined. As indicated earlier, Ihe placemenl
of these impacl treces is regular. occurring in Ihe
same lacalions relalive lo Ihe morphology of Ihe
dirrerenl bones.
1think il is dear Ihal Ihe combination of Ihe dis-
position and frequency of impacl scars, on both ar-
ticulalor ends and diaphysis fragments. coupled
with Ihe breakage morphology of Ihe arliculalor
ends, should correlate well with Ihe overall
marrow-bone-breakage laclics.
4. Human Modes of Bone Modificalan
Breakage Related to
Other Forma of
Bone Processing
Prevtous descrtptons extst regardng the use
of bones in renderins bone Srease (L. R Binford
1978b:157-163; Leechman 1951) and bone uice L.
R. Hinford 1978b:163-1651. In hoth processes the
bones carryng grease and fa! in the cells of cancel]-
ous tissue are pounded as lo increese the surtace araa
of tbe bone exposed. These fragmenta are then
botled. thereby rendertng off the greese. Spltting
and ctherwise pulvertatng articulalor ends is an in-
legral parl of this processinil. bul as we have seen il
is essentially never a consequence of marrow-bone
breakaile (with the occasional exception of Ihe prox
imal metapodials). I have previously discussed al
sorne length how one can idenlify lhe consequences
of Ihis behavior (see L. R. 8inford 1978b:t57-165).
Meo-to-Man Comparisons, or Altemative
Human Behevion
Given Ihe cenlral problem of diagnosing var-
iabilily in bones, one of the grealesl sources of var-
iabilily is mosl Iikely lo be Ihe differenl ways man
has developed for accomplishing similar things.
Clearly, if we are going lo compare man to ani-
mals, consideralion of Ihis polential variability is
I have by no means described 1311 the alternativa
laclics which the Nunamiut are capable of perforro-
ing. The slighl accommodations of lechnique lo the
differenl anatomical properties of Ihe marrow bones,
Iheir degree of articulalion. and Ihe degree of previ-
ous c1eaning, aU lead us lo expecl Ihal breaking Ihe
bones of larger animals. or bones from animals with
differenl analomical properties. would be accommo-
dated by modificalions in technique. In addition lo
the responsiveness of the Eskimo lo properties of Ihe
bones. Ihere is a correialed response to the situation
in which the activity is carried out. For inslance, ooe
informanl poinled out that using Ihe hand-held anvil
inside a winler honse during a meal might result in
bone fragmenls or articulalor ends going inlo Ihe fire
or someone's lap. or it would jusI be considerad
Mon-to-Man Comparisons, or Alternalive Human Benovtors
messy and "mpoltte." On the other hand, ustng a
hammer or an on-anvtl technique inside a house did
nol carry these rtsks. Similarly, men in hunling
stende repeatedly told me thet ustng the hammer-
on-bone stralegy was possible standing up end one
dd nol run the risk of having ends of bones, marrow,
and splinters fall to the ground and get drty. Men
also laughed and pclnted out that finding an anvil
and spreading a drop cloth was nct whal you did
while sitting around a re with friends! Finally, it
should be potnted out tha! there ts an nteracton
between ene sel of conditions of tecttce and another.
The amount of diserticulation and prevtous c1eaning
of bones is ap! lo be much less in a hunting camp or
sland. This is also where stand-up lechniques of
bone breakage are more appropriate. This means that
there will be sorne oorrelations belween technique
used and fracture patterning Ihal are nol Ihe direcl
result of lhe character of Ihe technique. bul instead
derive from Ihe characteristic coincidences belween
degrees of prior bone preparation and Ihe conditions
influencinil the type of lechnique used. This lype of
inlerection should encourage archaeologisls greatly,
sinee the slrenglh of such inleraetions for condition-
ingIhe resulling morphological properties ofbroken
bone should be slronil and primarily informative of
Ihe role of different Iypes of sites in the overall
roettlement-Iogistical system. This. if true. has slrong
potential for methodological development. These
descriptians will, 1hope. prompl the imsilination of
erchaeologisls to consider Ihe nature of lhe variables
man is responding to when behavinil in such a way
as to produce differenl morphological palterns in
broken bone. Doing Ihe lalter is Iheory building and
is Ihe basis for building a reliable sel of melhods for
giving meaning lo the archaeological record, It is
beilinninil lhe inleresting task of explllininil cultural
differences and similarilies.
ay liU) f 1ft,Ihe IIl!h g 4ltiV d 'p'ieAs
by Zierhut (1967) and 800nichsen (1973) p"........
, 'eb S dassffl .. "SsniFe f b b
One gets Ihe impression thal the CalHng Lake Cree
break bones by Ihe "mid-diaphysis smash lech
nique" everywhere and at aH times. in hunlinil
camps. in the house. snacking 00 marrow bone while
walking along a foresl Irail, while butchering ani-
mals. and so on. They do iI Ihal way nol because il
helps lo aceomplish a given lask in a given selting.
but beeause they are Galling Lake Cree! My guess is
that ths mpresston was structured by the nves-
getors and not by the Calhng Leke Cree. My guess
ts elso Ihal the Cree rarely eet marrow wlth meels in
the house. thet Ihey assumed the eonlexl from the
place where the quesons were asked. and their
demonstratons were more reminlscences of things
from the pest than actons daily performed. Whal
was demonslraled by the Cree was probably roughly
equivelent to the Eskimo's "bone on hand-held an-
vil" stretegy, generally performed in restdenttal
eamps. 1further guess that the marrow was tntended
lo be added lo a stew in small quanlities and not lo
yield a lerge quentty, as was most Eskmo women's
bone processing. Par! of my reasons for lhinking
Ihese thinils relates to my experiences as a young boy
in Appalachia. 1was lauilhl to break marrow bones
by my ilrandmolher in a manner similar to what
Zierhul (1967) describes for Ihe Calling Lake Cree,
except we invariably used an ax. You miilht call our
technique the "kindling wood technique" slthough
it meets 1311 Ihe criteria for Ihe "mid-diaphysis smash
technique" Ihal has so enamored Bonnichsen
(1979:54. Plale IV-ti. Slanley Soulh and I once dis-
cussed how we were taught lo break matrow bones
up in lhe eastern mounlains. Slanley said the frsl
time he was told to lale a big beef bone oul and
"break il up" because "Unde so-and-so was coming
who really Iiked bone marrow" was a real learning
experience. Slan said he went out to the woodshed
and put Ihe big bone down wilh each end resting on
a split section of 10il and gava a mighly swing with
lhe butt end of lhe ex. The bone went flying off inlo
the woodpile. He tried ailain and the same thing
happened. He then turned the ax around and hit Ihe
bone rnidshaft wilh the bit of the a", and the bone
broke nicely in half. bul one end new up and hit the
roof of Ihe woodshed. Although my experiences
were largely the same. 1 had been told by rny
grandmolher always lo hit Ihe bone in Ihe rniddle
wilh Ihe bit o Ihe "old dull ax out by Ihe stump
hamess." This was an ax used for culting rools when
stumps were being pulled by amuJe leam. It was
always well polished from having been driven inlo
the soil and fairly dull from having hil so many
slones. 1had also been lold not lo hit il as hard as a
piece of firewood "because it mighl fly up and hit
you in Ihe head." Chmrly 1had no knowledile of bone
analomy: an)' bone given us was trealed in Ihe sarne
fashion. Similarly we cerlainly had no knowledge of

fracture mechanics ocarry understanding of Ihe rela-
tionships belween bone anatomy, tmpactlccatton. oc
strength of loading, as is illustrated by Eskimo be-
havior. We jusi learned that ir we hit it too hard it
might "fly up and hit U5 in the heed." In short. we
used our knowledge of chopplng kindling wood to
fracture bones. Although I ha ve ergued eganst Ihe
analogy with ltthtcs as a guide lo differentiating be-
tween animal and humanly broken bones. c1early the
knowledge used by tbe Eskimo wes more akin lo Ihe
knowtedge needed in lithics reducon and 1001
manufacture than to that needed in kindling cutting
with un 8Jl;!
The kindling-wood strategy Ihat I learned as a
youog boy, which Zierhul (19671 and 80nnichsen
(1973) observad amnog Ihe Calling Lake Cree. and
which 1 observed amoog the Navajo (see L. R. Bin-
foro 8nd J. B. Berlram 1977:94), appears to bea prep-
Ulltion tachnique altlociated- with- nrlatively. aect.n-
tary peoples using boiHng ss 'en IIIJ1'1$f exclusive
technique of food pntparation: It appears lo he
modeled on Ihe mechanics of cutting kindling wood
with an ax rather Ihan on an understanding of the
fracture mechanics of the bone ilself and Ihe rela
lionships of impact to bone morphology. Both Ihe
apparenl model for the lechnique and the contextual
features common to Ihe Cree, Navajo, and Ap-
palachia suggest Ihal this is a very poor ethnog-
raphic analagy fOl modeling the behavior of hig-
game hunters, men of the Pleistocene. or foc distin-
guishing the products of men from the products of
I am nol suggesting Ihat we substitute Ihe almosl
certainly incomplete descriptions (rom Ihe Nuna-
miul Eskimo as our generalized configurational
model of marrow cracking. Instead I am suggesling
that we attempl to understand why one group of
people does Ihings one way and another does similar
Ihings in differenl ways. In short, explaining cultural
differences and similarities is our iob. If we cannol
do this we cannot develop means for reBably giving
meaning lo our observalions. For instance, in a very
important seminal sludy of marrow cracking. Ihe
fragmenl morphology at the famous siles of Stac ['..arr
and Kongenmosen was compared wilh Ihose from
Iwo olher wel1 known Mesolilhic sites from norlhern
4. Human Modes of Bone Modi!kalJon
Europe [Noe-Nygaard 1977). The author cleerly
demonstretes thal the morphology of fragments re-
maining on the Slar Carr si te s similar lo thet et
Kongenmosen. end both differ from the fragmenr
morphologv characleristie of the Praestelyngen and
Muldbjerg J si tes. At Star Cerr long bones tended lo
he broken in midshaft wilh most of the dleph ysis
remaining attached to the articulator ende. By way of
contrast. at the sites of Preestelvngen and Muldbjerg
I the bones seem lo have been crecked for marrow
usng a technique thet directs Ihe irnpact in the area
of the artculator ends. similar to that generally de-
scrtbed bere for the Eskimo. In Iact. comparson of
the iIluslrations {Noe-Nygaard 1977: Figure 8) and
Ihe specimens illustraled here (see Figure 4.53]
strongly suggests Ihal similar techniques were used
by Ihe Eskima in residential eamps and Ihe occup-
ants oC Praestelyngen and Muldbjerg 1. On Ihe other
hand, Ihe materials iIIustraled from Slar CaIT evi-
dence hreaks similar lo those produced in the hutch-
ering of Crozen careasses ralher than those produced
in marrow recovery. The low frequencies of femurs
(Noe-Nygaard 1977:222} are consistent with a kili
bulchering location or a {ield camp (see L. R. Binford
1978hl. Similarly, Ihe low Irequencies 01 the prox-
imal humerus relative lo the distal humerus plus Ihe
reported presence of femoral cylinders strongly
suggesl lhal Ihe remains from Slar Carr were ravaged
by predalor-scavengers. The tolality of Ihe dif-
ferenees noted belween lhe four sites strongly
suggests a majar contrast hetween residential loea-
tions and special-purpose sites related lo hunting
10gistics. These approaches lo underslanding Ihe dU-
ferences contrast wilh the more Iraditional view:
Differem;:es in msow fracturing technique between
various cultureSrefleel different methods of preparing
food and may therefme rened differences in culnnal
level. such as aceramic a ~ opposed to ceramic cultures.
JI may thus be possible lo use Ihe character of Done
splilling as an ecoslratiRrltphir.al 1001 INoe-Nygaard
Clearly we need lar more delailed pattem-
recognilion .'ltudies. designed to elucidale Iracture
palterning, of archaeological bone assemblages. In
sorne cases we can oblain "controls." in Ihat Ihe
ComparingMon and Becst
speces can be reliably identfled end any currela-
tons with body stze or with bone density can be
Investlgated. Stnularly. where we have good reeson
to place sorne faith in our functional interpretations
01 sttee. sueh as blson jumps and kili sttes versus
residential locatons. or processing reas eseocated
with bison kflls, we may begn to determine how
sensitive fracture patteming is to site functional df-
ferences. Similarly, if we can gan control ovar "sea-
son of occupancy," we can begn to nvesttgate the
degree lo which butchertng strategy may be altered
lo "bone breakage" versus "[otnt cutting" during
seasons of extreme cold. We have not even begun lo
investigate the potentia! information aboul Ihe past
baing earried in Ihe archaeologicel record by sueh
mundane and unspectacular Ihings as broken bones.
As inleresting as such implicetions for fulure
study may be, our real interest hece is with criteria
that mighl dislinguish bones broken by meo Irom
those fractured by nonhuman predator-scavengers.
Comparing Man and Beast
I have thus lar developed a number of conlrasts,
such as eul marks from lools versus !ooth-scoring
marks. Impact fracturing from the use of tools in
bone breaking has been conlrasted lo Ihe gradual de-
struction of 50ft parls by the progressive weakening
of hones and their col1apse under animal persis-
lence. Slanding behind these conlrasls is the recog-
nition tha! animals and man may well dismember a
skelelon differently because each aels under dif-
ferenl mechanical conslraints.
There is Jittle chance Ihan an observer of madi
fied bone would confuse cul marks inflieted during
dismembering or filleling by man using lools with
Ihe action of animals. It was noted Ihal cut marks are
coneentrated on arlieulator surfaces and are rela-
tively rare as t r a n s v e r ~ e marks on long bones excepl
whem encircling marks may betray skinning activi-
lies. Jt was suggesled thal cut marks from slone tools
are mosl commooly made with a sawing molion re-
sulting in short and frcquenlly mulliple bul roughly
parallel marka. Such marks are generally cherac-
lerized by an open cross secton. Another cherec-
tenstc of cut marks derived Irom the use of stone
tools is that they rarely follow the contours of the
bone on which they appear. That ts. the cut does not
show equal pressure in depresaons and elong prurn-
inenl ridges or across the are of a cylinder. In al-
rnost all ways. the marka of animals' teeth are sorne-
what different. They commonly follow the contours
of the bone's surfece. as is well illuslrated in the
specimen shown in beautful color by National Geo-
grophic (Canhy 1979:356) cted as evidence of rnan's
presence al the site of Tagua Tagua in Chile. The
caption reads, "Deliberately fraclured and engraved
with parallel strokes. a horse leg bone was found in a
natural trap. . This is almost a type specimen of
looth seoring by gnawing animals! Tooth marks may
frequently lake the form of a depressed or rnashed
Hnes, as appears to be Ihe case on the bone from
Chile. On many of Ihe wolf specimens, Ihe loolh
rnark under magnificalion appears as a "cracked"
5urface scar rather Ihan as a cut or incision in Ihe
bone as is Ihe case particularly with metal tools.
Claims for the use 01 lools should be supported hy
the cilation of rnarks produced by lools.l Admit-
tedly, distinguishing belwefln toolh scoring and cut
marks is sometimes difficult. but the attempt rnust be
made. For instanee, increases in the amounl and
densily of scoring and pitting are noted around areas
thal have been partialIy destroyed or furrowed, par-
licularly 8round long-bone shafts when the ar-
lieulalar end has been partially deslroyed by gnaw-
ing. This association of scoring wilh paltems 01 de-
struction is not lo be expecled when roan dismem
bers an animal with lools. More cornmonly the
modifications produced by Rnimals have been inler-
1) musl emphasizB lhst I am speaking of a subslanlial
populalion of animals and nol a single animaL) am sIso
"hedging" on the size of lhe populalion. for if one exdurles
Ihe mandible lhe highesl proporlion of marked bones is
30..... 0% of Ihose presenl in Ihe Mouslerian sample .'lee
Table4.03). On lheolher hand, around 50%weremarked in
somedasses represenled in lhe modern Nunamiul maleriltl
(see T;ble4.02) Obviously in small samples we could ex"
p('cl some examples ofhumanly butc:hered animals exhibil-
in!! rew ir any marked bones.
peeted as either intenttonal modification oCbone for
use as tools nr attritional modcetton from the
bone's having been used as a too1. The use and prod-
uction of bcne tools haya nol been discussed here,
for eeveral reasons. Most importan! is the fact thet
quite literally this s frequently the "bone of contn-
tcn," the tnterpretetton of uncertenty. If 1Receptad
everything that has been clted as a tool sud pro-
ceeded to describe it, then elmost al1 the
documentad modcetons produced by enimals
would haya been descnbed. Do Ihe albar hend, ir 1
had demanded tha same actualistic controle. the bear
end the footprint seen together, as the baste for de-
scrbng bone tools "cheractensttcelly produced" by
human egents. then 1 would face the unlikely uni-
{armitarlan assumption Ihat conlemporary patterns
of manufacture and use were Ihe same as in tha pasl!
We are not lalking aboul products generaled because
of "intrinsic" mechanical and behavioral constraints
or Iimitations on the way a bone can be modified, as
is likely to be the case of animal modificalion.
Human tool design and use are responsive lo situa
tion, mechanical needs, and the character of know-
ledge, equipment, and raw materials available. Tool
production is responsive to tool need and recognized
oplions, among other things.
We know Ihat the range of tool needs, the charac-
ter of knowledge, available equipment, and accessi-
ble raw materials are certalnly different and repre-
sentative of a different range o variability today than
would hava been the case at different periods in
man's pas\. Wilhoul knowing the things we seek to
know about Ihe pasl. we could nevar guess whal con
temporary experiences mighl be directly relevant to
eilber a specific archBeological case or all tbe past
taken in a generic sense. We connot ossert thot we
understond the very phenomenon we seek to under-
stand in arder to achieve a method for its study!
Argumenls about bone 1001 use and manufacture
musl be arguments from elimination. where alterna
tive explanations are syslematically eliminated
Ihrough the use of a robusl and informed melhodol
ogy. Even after we have narrowed the idenlity o the
causal agenl to manoIhe question Ihen remains as to
what man did wilh the item and why he made il in
Ihe manner observed. For instance. I have made fre-
quent reference lo Ihe facl Ihal animals produced
modified bones Ihat were to me indislinguishable
from whal have been called fleshers by Frisan lt97BI
.- CA--.. f"r
4. Human Mcdesof Bane Modification
and others. In Ihis case I am questontng the identity
o the causal egent. On the other hand. Frisan has
repaaledly noted a pattern of modtcaton on the
fractured ends of long bones at eeveral bteon kili
butchering sites. He has nterpreted theee as chop-
pers used expediently in Ihe butchertng process (see
Frison 1974: 51-56, 1978: 301-316}. Lagree thet many
of the uems iIlustraled by Frison have no direct
enalogues in enmal-modttted assembleges. Vague
similarities can sometimes be found for sorne bones
with dense shafts such as the distal tibia, hui 1 have
been unable to nd anything similar lo the proximal
tibia and proximal femur choppers iIIustraled by Fri-
My experiences with butchering and expedent
marrow eating by butchers while engeged in work,
as well as Iunius Bird's experiences in Palagonia as
reported by Flannery. {Hole el al. 1969:291J suggesl
Ihat the same patterns of modification Frison ob-
served on his choppers might derive from striking
the broken ends of fraclured bones clown on hard
surfaces during marrow removal.
This is the type of problem to which experiments
can he direcled. For nstance, in the case of Frison's
post hoc interpretative model regarding chopper
use. abone could be used in the manner envisioned
by Frison. Similarly, an identical bone could he
treated in the manner described for marrow crack-
ing. The consequences o the Iwo separate acts could
be compared, both could be compared to the ar-
chaeological bone, and one of the alternatives could.
it is hoped, be eliminated.
There is stiU anolher research approach. the ap
proach Ihal stresses the "integration of imowledgs"
as a criteria of plausbiJjty. For instance, as long aS
there was reason lo believe Ihat ancient men on the
Plains muscle stripped animals as a normal butcher-
ing stralegy. then Ihe use and presenee of choppers
were inlernally consistenl with tha facts cited as evi-
dence for muscle slripping. I have seriously ques-
lioned the degree lo which the (acts cited in support
of the muscle-stripping argument are referable to
human action. Mosl of the forms of hone destruction
inlerpreted as evidence of musde slripping are
seemingly indistinguishable from animal-ravaged
faunal material. If this is Ihe case, then there should
be no necessary correlation between choppers and
Ihe alleged evidence of musde stripping. Carrying
oul such comparalive studies would almost certainly
Morphologr of Bone Breakoge
result in the recognton of previously unrecogntzed
patterning in the ercheeclogtcal record. These pet-
terns then would present us wilh new challenges
and potentel sources of new nformeton regarding
the pest. Knowledge gained through such prompted
reseerch> than would mcreese the strength of our
methodology and in tum our scenca would grow in
both knowledge and sophistication.
The problem of surficial modification, although
ccmplcsted. is relatvely minor comparad lo the
problem of argumente thet cite structurel modifica-
lions and breakage petterns as causally referable lo
human agente. Most cletms for human causaly
have reference to overall morphology or the presence
of obvious postmortem modificalions on bones, such
as polish or abrasions considered indicative of use.
It sooms lo me thal hefore interpretations are offered
for sucb modifications one must attempt lo assess
the relativa role of different agents as general con-
tribulors lo Iha populations ofbones from which such
"inleresling" items mighl be singled oul. Arguing
for human modification of a particularly inleresling
ilam abstracted from a populalion of otherwise ex-
c1usively animal-modified bone is not very convinc-
ing. This is critera ofthe "integral ion of knowledge"
somewhat reversed and used as a procedural sug-
gesUon for research. Are thete nol properties of
populations Ihal mighl he evalualed lo provide the
conlext for looldng at the charactaristics of partic-
ular items? The following section presents a com-
pilalion of sorne properties of humanly modified
faunal assemblages lhal contrast in a generic sense
with properties of assemblages produced by ani-
2What 1havecalled here prompledresearch issimilar to
what RichardGould calls the orgumentby anomoly. 1rec-
ognize thal when we are surprised Of observeunexpected
things. we should investigale(seeL. R. Binford1972b:115].
InIhis 1 concur with Gould. Thlll this in IIny wayiSII more
epistomologically "convincing epproech then !he of
analogy" (R. A. Gould lCl80:138) is sheer An er-
gumenl from analogy is 11 form of inductive argumen!.
Gould's argumenl byonomalyis nol a form af Ilrgumenl but
a guide lo asking inleresling QlIestions. The answers to
such queslions maybegiven !hruughIIfgUmenls from anal-
ogy. which is exar:tly whal (;oulJ in his eomple of
"anomlllous" rllw material usage al PUlllUljorplI R A.
Gould HIBO: 153-1SfiJ. despite his
Morphology o Bone Breakage
The degree to whch bones have been chewed and
fed upon conditions the character of the survvng
faunal Inventory. However, tbe pettem of bone
gnawing is redundant and reeults in a characteristic
pettem of bone destructon. This pettem was rst
described during rny AJaska eldwcrk by Dan Wit-
ter, who was given the task of studying the bcnes
remenng in eeveral well-documented dog yarda in
Anakluvuk village. Later. when I worked over my
documentation of wolf ktlls. it beceme olear that the
same types of destruction were presento
There is a fairly regular sequen_ce lo Ihe destruc-
tion of bones by both wolves and doss. The redun-
ailncy in the chewing strategy results in there being
very distinctive pattems to surviving long.bone
parts. A general c1assifcation of Ihes e parts was
worked out by Dan Witter, and his terminology will
be adopled here.
1. ArticuJotor end (Figure 4.56, spedmens on
left side): the articulator end when Ihera is a
broken diaphysis just below the epiphysis of a
bone. That is, Ihe articulalor end has been bro.
ken offimmedialely below the "head" or in the
neck of lhe bone, with HUle attached
2. End plus shonk (Figure 4.56, middle speci.
men): Ihe articul8tor snd with aHached
diaphysis complete lo approximately halfway
between Ihe lwo ends.
3. End plus shaft (Figure 4.56. specimens on
righl side}: 8n articulalor end with tha aHached
complete diaphysis ranging from half the
lenglh of the hone as far as the neck at the
opposile end. The bone is essentially complete
except fer Ihe opposite end.
4. Cylinder (Fisure 4.57): the central parl of long
bones or segments thereof. minus Ihe ar-
ticulotor ends. The behavior that produces Ihis
Iype of parl is simply Ihe altemate gnawing
away of Ihe articulator ends leaving segments
oflhe long-bone shaft remaining as a cylinder.
It should be noled that as Ihe bone i8 reduced by
gnawing Ihere is a correlaled destruction of
anatomy. Abone preserved as a dislal articulalor end
plus shaft (Figure 4.56. righl sida) is rartlly as "de-
M"fph",,g.l' (JJ Honr- Ilruukug,
1 4. Humnn Modt->s uf HOJW Modifinrfiun
tlCI TRt: 4.51j, ../ difl"n'lll JI! Im-1I1ww ..nI: I""w._.
t'If;(IRE4.51. Cdiflllr'/","_I /iU#'Ul,OI' produ(;/ "! animal
stroyed" un the distal ond as ls a distal aruculator
end when both the shaft anrl shank have bonn re-
moved. Tl:o ,! I hi!ifj!l9tlCJ f" p *6 ll'iUllJ.

'l;W:iz: O. Tll..i1fl 4.07 summarizes
data from the wolf Io.ills. three tlf}fJ, .vanl samples. and
lWfI IHm;pssed by man as \0 Ihe
frequcncie.s uf idenlifiable parls represent-
ing difieren! slages of Uf <IHritiOll. T,-",""-
tic of ROaw.illS aPirnaJSc.
aSSOGiR1t,d wilh breakage o[ OOJle&--by
mm foro mllrnU\l .cxtra,ctioll.
1 Animill .111 attriliullal Te-
sullng in Ihe progn>ssivt: dell'liotl [)f cerlain parls
fmm 11](' nf b01Wfor 1111'
erv 01" marrow by ruan is il pr(l{:l'SS that n-suls in
morphologn.al ch,Hlges but nct necessarfy ru-
sult in aurttion.
2. Animal gnawing ts ti sequeritial prot.ess rasult-
ing in a progres-uve development 01' deslrllclivp pat.
terniug tha t has 1\ signalure qualv cormiatnd with
[hu patter-n of breakage Exploitafion of bunes for
mnrrow bv mnn is a diruct proccss su eh Ihil( tllHre i::.
no necnssarv correlaon hetwccn the morphotogv uf
breakage and the pattem of survivorship in hum-
Given too!s. man is able lago dlrectly lo thu turget
of explottauon. If skllod. he first cleans lh" bune so
control of oercussio (smantained {seoL R. Binford
197Hb:152-157). and Ihen usmg a hammer or anvil
breaks the bcne generally [ust beiow the head or ar-
uculator end nf the bone. RfJgardless of Ihe varying
skills involved and the Iact t hat tbere are different
ways of addressing the bcne wlth percussion tools.
tbe pctru bf'ing made is not altered: breakaga is ac-
complished by directly. and is nct dependeut
on weakenng the bone thruugh the successive de-
strur-tion of parts. as is the case with gnawtng.
Figure 4.58 illustratss the relntionship between
percfJnlage Irequencies of broken bone parta as tabu-
laled in Tahle 4.07. columns 2 and 14, for Ihe wolf
kills disp laved againslllw sample from Ihe Kakinya
'lile. columns 12 ami 24. There are many examplcs uf
hih percentnge values rol' arliculetor ends in lhe
Kaknva sarnp!r- ano correspondf ng low values in the
wolf kill sample. Conversely. lhme are high percent-
age valas or arficulntor euds with aHached shafts
and shanks in the wolf kili populafion and corre-
sponding low vales in the Kakinva sample. The
basir: rlislribulion is uf an indirecll'oissoll form with
a separale dump of arliculalor ends Ihal show mod-
erale lo high valum; in Ihe wolf kili sample. These are
bolh ends oflhe fmnur. Ihe proximal metatarsal. the
proximal libia. Ihe dislal humerus. amI Ihe distal
radio-cubiluf'l. '1'0 lhe 1.J<ltterning sOlnewhal dif-
ferent!y. high heqllendes of arHculator ends wilh
alta\:hed shafls and shanks are c\J.araclerislic of
lI11imal-gnaw(1(j assemblages. Isolaled arti<;ulator
may Dccur in h(llh I.VjJes \lf asspmhlages bul
arl' largnl\' lhe or dominant fOflllS in as-
cracknd lor marruw b\ mall.

4. Human Modes of Bcne Modificolion MorphoJo&Y of Bone Breakage
TARLE4.07 TARLE4.07-Continued
VariatloR In BOlleBreaka8e larAn/mols and HumanDel'ived Assemblaes
Dog yarda Human-dcrived assemblages
D08 yarda
Human-derived assemblaglOl5
Wolf kill& Sod 2 Sod 3 Sod r Anavik stand Kakinyasita
Wolf kills Sod 2 SO<! 1 Sod J Alllllvik stand Kakinya sue
Anetomcel par! (1) (2) (1) i') (5) (51 1') (51 (9) (ID) (lt) (12) Analomical pert 113) 114) 115) 116) (17) (18) {19) 120) 121) \22) 123) \24)
Humerus Femur
Proximal unly 1.0 .40 1.5 .50 .5 1.00 4.0 1.00 Proximal only 15 41 10 .67 1.5 1.00 1.0 1.00 1.0 1.00
'5 ...
Proximal + ebenk .5 .20 Proximnl + shank 5 .H
.5 .06
PlOximal + shaft Proximal + shafl 5 14
Complete 1.0 .40 2.0 1.00 1.5 .50 COffiJlh!te 10 .29 5 31
Total 1.5 2.0 O 3.0 .5 4.0 Total 3.5 1.5 15 1.0 1.0 9.0
Distalonly .5 .05 .5 10 2.0 .57 40 .19 5 1.00 5.5 .83 Distalonly 50 .71 5 t4 .5 1,00 1.0 .e, r.o 1.00 '.5 .zs
utslaJ + shank 2.5 .26 1.U 20 5 .H '.U .43 .5 .0' Distal + shank t.n 14 1.5 43
2.0 .21
Di!ltal + shafl 5.' .58 1.' 30 10 .2' e.e .31 Distal + shaft 1.0 .14 .5 .14 .s .33
Complete 1.0 .11 20 .40 10 .01 Complete 1.0 .14 5 14
Total 9.5 5.0 35 n.n .5 6.0 Tolal 1.0 3.5 .5 15 1.0
Hadio-cubitus Tihio
Proximal only 5 .06 .5 .11 5 .t4 1.5 .0' 10 .86 1.5 .25 Proximal only 1.5 .30 5 1.00 2.5 .71 3.0 1.00 '.0 1.00
Proximal + ehenk 1.5 .1' 1.0 22 5 .14 5.0 .29 1.5 .25
Proximal + shank 5 10
Proximal + sha '.0 .50 .5 11 5 .t4 5.5 .32 Proximal + shaft
COmplele '.0 .25 25 .56 20 .57 5.0 .2' 5 .33 Complete ,O 60 5 1.00 1.0 .29
Total 6.0 4.5 3.:; 17.0 '.5 6.0 TUlllJ 5U .5 .5 3.5
'.0 1.0
Dlstalonly 15 .33
4.0 .89 5.0 1.00 Distalonly
15 .25 15 .11 3.0 .88 4.5 .90
Distal + shlnk .5 .11 1.0 25 1.0 .H Distal + shenk .5 05 1.5 25
'.0 .22 .5 .14 .5 .10
Distal + shaft 1.0 .22 .5 ra 10 .H Distal + shaft '.5 66 2.5 42 1.0 1.00 '.0 5'
Complete 2.0 44 2.5 63 2.0 1.00 5.0 71 .5 .11 Complete 3.0 .21 .5 06 1.0 ..,
Total 4.5 40 20 1.0 4S 5.0 Total 11.0 60 1.0 13.5 3.5 5.0
Metacarpal Melalarsal
Proximal only .5 .0' .5 .20
1.0 .50
Proximal only 5.0 .5' 3.0 67 .5 .50 3.5 .70 3.0 1.00 2.5 .63
Proximal + shank 10 .18 5 .50 1.5 .30 Proximal + ,hank
Proximal + shaft 15 .27 .5 20 15 30 Proximal " sheft .5 .DO .5 .50
Complete 1.5 .45 1.5 80 5 50 2.0 .40 2.5 1.00 10 .50
'.0 35 10 .33 15 .30 15 .38
Total '.5 2.5 10 5.0 25 2.0 Tolal 6S 30 10 5.0 30 40
Distalonly 15 .21 1.0 .33 1.5 .38 .5 .33
5.0 1.00 1.5 .38
Distal + shank 1.0 .16 Distal ... shank 1.0 22 r o .33 1.0 .24
Distal + sbaft 5 .09 Distal + shafl 5 11
.5 .H
Complete a .45 15 1.00 .5 1.00 2.0 .86 2.5 .63 1.0 .57
Complete 3.0 67 1.0 J.OO r.e .50 1.5 38
Tolal '.5 1.5 .5 3.0 4.5 15
Total 45 10 O 3.0 5.0 4.0
centege o (he maxtmum number of artculetor ends A tota!l
different stuetton holds for sttes where
number of erttculeto- ends present 00 thesite when
!he 4.57):
that could have been present but were in fact missing liumaos were res onsi a for a essem ege. correcttons were made for the number of complete
nafa on cylinderlt and ... pl inters are gtven in Chapter (24 in the case Df Idlls o single eatmals minus tbe cept in CRse-; where therc waa menu ecture o
oue bones and the number nf shafts and cyl nders (see L.
5, Table 5.01. Figure 4.59 tllustretes the relenonehtp
nurnber present divided by 24). There is clearly a grease or bona [uce (Intentonal destrucnon of ar-
R. Binfcrd 1976b:157, 464, Table 9.1).
between the numhet o bone spttnters pet MNI from
postttve linear relationshlp between the number of
ticulator endsj. Ihere wes a fairly constent relatton-
Another Interestmg relationship s demonstrated
tbe wolf kills (dscussed in Chepter 5) and Ihe pero
spltnters and the ertculetor ends that were msstng.
ship between the number of bone splinters end the
in Figure 4.60, where the number of splinters is dis-
-._.. _--

4. Human Mod!'s nf Ronr Mndifiration 5ummory 177
ro M
TABLE 4.01, COLS 'l ANO 24
AGURE 4.58. Relatlonship between breokase /I-equen-
c:ie. al the Kaldnya slre (human} and on Ihe AJask"n wolf
Idlls (animal).
FIGURE 4.59. Relationshjp betwoon number 01 bone
spllnrers per MNl and lhe percenlage 01 articulotor ends
potentlally present thot were in lod (Numbers
relate to in ToMe 5.01.)
In Chapler 1 J suggested Ihat in the earliest days
o archaeologrcal reseerch the paradigmatic view
constdered there te be Iwo "creators"--God, the
molder o neture. and manothe molder of culture and
cwtzauon. Eorly argumenta regarding evldence for
men's antiquity developed the position that tha pet-
temed modifications observad on sorne slones could
not have heen produced "in natura" and had there-
fore to be a consequence of the hand of mano Incor-
rect judgments were made in the early days uf ar-
chaeologrcal research as lo what forma of patterning
nature was incapable of productng. This ignorance
of nalure led to the spurious esetgnment of many
things to the hand of mano In most cases this seme
siluation can be vewed as responsible for neccurate
nterpretatons of modified bones as tools, spiral
fractures as only produced by men. and many other
such inaccurate or at best ambiguous assignments of
meaning to bone modificalions. In Chapter 2 1
suggesled that as far as middla-range research is
concerned, we are seeking lo isolale causal relalion-
ships belween energy sources and modifications in
malerial things. If we can galn a causal undersland
ing of certain relalionships. then Ihere is established
a relotionship of necessity between the cause and
the effect. When such a necessary relationship can be
established then Ihe inference of the cause when ef-
fecls are observed IS particularty secure,
When the early researchers into man's antiquity
addressed the subjecl of stone tools, the comparisons
were belween one agent-hominids-and ather in-
animale nalural forces Ihal might alter slones. The
difference belween palterned goal.directed behavior
of man and the accumulation of modifications thal
were the resull of non-goal-direcled sequendng ren-
dered Ihe recognition of stone lools largely unam-
biguous. The crilerion of 'manufaclured lo stan-
dards" or repetitively "produced in the same way"
served for recognizing human products in slone.
A relationship of necessily was implied between
complex sequential modification o matler to a pal-
tern and man as Ihe actor. This was largely correcl
for stone 1001s, since man is essenlially unique in his
use of stone as a raw malerial for the produclion of
lools. Sinee Ihis relalionship belwcen bear anrl foot
x = lA - (ZE + GlJ(y)
the more spltnters remain. For instance, uslng the
methods found so successful for estlmating the ex-
pected number of bone spltnters present on asile
given a count of the articulator ends from humanly
broken bones. we obtain sorne surprising results .
The total nurnber o artculator ends observed in the
combtned semple of wolf kills was 145, of whch 50
were artculator ends 00 complete bones. Thera wcre
thus 95 broken artculator ends. Using this figure in
the formula worked out previously (L. R. Binford
lo short, no splinlers are given Ihe surviv-
og population, yel we observed 1119! This Ilus-
trales nicely Ihal the splinlers are related primarily
lo whal is no longer presenl in Ihe animal-gnawed
assemblage, whereas in Ihe humanl}' borken as
semblage the splinters are relaled lo wha! remains as
uodestroyed articulalor enos.
I feel confidenl Ihal I have demonstraled that in
an animal.gnawed Iissemblage we may partition
Ihe assemblage into hasieally two componen's-
splinters and cylinders-which vary in thtJr fre-
quencies primarily with regaro wilh what is no
/onger presenl in Ihe form of articulator ends or
complete bones. Complete bones vary positively
with shals and arliculalor ends represenlin;; the
elemcnls or survivurs not by Ihe gnawing
x = expected number of bone sphnters
A = number of broken erttculator ends
E = number of cylinders (Table 5.01)
G= number of ends plus shaft (Tahle 4.07)
y = constant o relevance. in this case Z.85 or the
mean number of splinlers produced per arliculalor
eod when hones are broken for marrow
Substitutiog io this equation the foregoing dala,
we oblain Ihe following results:
x = 195 - (15Z - 43)] (2.85)
= (-100) (2.R5)
= -285.0 splinlers
s o I e 70 B o
.' "
", I
"l' ".
JO ", ...
20 '.@\.
'0 '. n,

O ----,---.20 :lO 40 60
C1l.IND[IlS ...NI
to be produced by gnawing animals; hence breakage
ncreeses as the animal progresses in the reduction
of abone to a cylinder. Figure 4.61 lllustrates the
converse of this relalionship. namely Ihal the
number of cylinders per MNI decreases as the prop-
ortion of Ihe arliculalor ends represenled by com-
plete bones increases. The more the destruction. the
more cylinders and the fewer complete bones there
are. In lurn, the more the deslruction of bones per se,
-0'1 .;

20 .' .'4
, .'
'O 20 JO 40 60 70 80 'tD 'DO

FIGURE 4.61. Relotionship belween the percenlage of
total artjeulator eRds reprP.senled by completp. btJnli's(Ind
the numbcr 01 cylinders per MNI. (Numbcrs relate ta (;u/
in Tablc 5.rn,)
FIGURE 4.60. ReJotlonship between number 01 bane
spJlnters ond bone eyUnders per MNl. (Numbers reJatelo
columns In TaMe 5.01.}

.' ,:;..'"
'''' .. ::;
""'\ l' Of

, J'."

.' " . ..
"'.l @,

'u 20 -lO 40--;0-60 70 80 '10 100
% UTICUlA.. n.o:; "":;S'NG
! 10

! ro
"j: 60

; ,

<el :::O,.

played eganst the numbers of cylinders per MNI
[taken from Table 5.01). It is cleer that a strong pos-
tive linear relationship is indicated up lo ao approx-
mate value oC55 splinters per MNI. Al tha! poin! Ihe
trend is reversed and the number oC cylinders is re-
dueed in ao apparent linear fashion with further in-
creases in the numbers oC splinters. This simply re-
flects tha! cylinders are Olle oC Ihe "lasl slage" forms
print was essentially one of necessity DI uniqueness
in natura, we nave been Iergely successful inferriog
actions by men. (Ihe beerj. from his stone tool prod-
ucts {the footprinl).
A tacttcel analogy with lithic studes ts probebly
the basis for mcst of out neccurate interpretations al
bone tools. As Illustreted in Chapter 3 nonhominid
animals regularly modify bcnes: they manipulate
both bones and prey antmels in ways that result in
regular, eequenelly generated. and pattemed
modifications. These mcdtcaons extend from in-
dividual bonos through units 01 larger anatomical
organization to complete faunal populations. Man is
nol relatad lo patterned sequential modificatioo of
fauna! material in a unique or necessary way, as is
Ibe case with lithic materials. Failure to take into
consideratioo alternative causal sources of patterned
modificatioos noted 00 bones further exacerbated
the probIem rooted io the assumption Ihal a1l things
found in association with stone tools couId also be
attribuled lo Ihe hand of mano 5uch an argumenl
seemed particularly secure when those aS50ciated
things exhibited pattern rnodifications, which were
taken to indicate "manufaclurad to standards" or
"produced in a standard way."
In Chapter 3 1 spenl sorne time illustrating how
many patterned properties of modified bones, citad
by olhers as evidence for human modificalion or use,
were in fact regularJy produced by animBls. 1 at
templed to Brgue that many of the modifications
were nacessary consequences of the anatomy of an
animal's jBw, such Ihat if modificalions wera lo be
made Ihey would inevitably be made in cartaln ways.
1 soughl to iIlustrate Ihal an animal tearing apart
a skeleton was limited to a certain sequence and lo
e given strategy by virtue of the constraints on Ihe
use of jaws as the animal's basic mechanical device
for carrying out work on bones and carcasses. Thal
is, 1 have attempted to at least suggesl sorne of Ihe
uDique. or necessary, linkages between gnawing ani
mals and their products.
In Ihis chapter, I have explored sorne of our
knowledge regarding the ways man manipulates and
modifies bones. A\though our knowledge is surpris-
ingly sparse, seem in arder.
Perhaps the most important point to be ma'{'f(;1s
that abone may take on properUes from a number of
sequentially independent acts. Treating all Ihe
4. Human Modes of Bone Modification
morphologtcal properties as if they were referable lo
a single hstortcal episode or behevtor s a majar er-
ror. We need informed and detetled dtagnosttc
criteria for recognizing properties that derved from
independent acts in the pest. This should be olear
with regard te separating consequences of animal
bebavor from thet of humana but it is equally trnpor-
tant to be able lo recognize derivativas o derent
human ectons.
Of equal mportence to the recognlton o mubt-
pie causes s the realization thet similar properttee
such as pattems of bone breekage may be produced
in a number of differenl contexts. Breekege may
occur in the cootexl of bulchering, in Ihe contexl of
field snacking for marrow. in the context o residen-
tial processing o marrow bones, and so forth. Trelll
iog all breakage as referable lo butchering. or
marrow cracking, or any other single behavioral con-
lext by convention is a major error. lo the absence of
lruely diagnostie means for recognizing the conse-
quences of one context of breakage from another. in-
terpratation by wnvenlion will certainly lead lo
myth making.
In terms of dstinguishing properties generated
by animals versus charll.cterislics referll.ble to hom
inid lI.ctions, a common stralegy has been to seek
characteristics thll.t might be generically referable to
"human" behavior. We have seen Ihis with the
daims o uniqueness for Ihe crackandtwist method
of bone breakll.ge. Ihe longitudinal breakage o long
bones, spiral fracture. pressure "flaking" of bones,
polishing, striatioos, and the Iike, whieh have all
been considered by sorne lo be diagnostic of hominid
behavior. Can we cite such diagnosHc properties
based on the studies and knowledge reported here as
necessary consequences of hominids achng upon
We might ask initiBlly whether there are any gen-
eral poiots of contrast between what we know of
human behavior and whal we know of animal be-
havior. In this we musl answer lhal Ihere are pro-
vocative differences.
Man ls not limited lo Ihe physical manipulalion
of maleriallhings with a single mechanical device,
jsws. Man conceived in very simple terms as a tool
user can manipulate physicallhings with a variety af
mechanical devices Ilnd in lenos of a number of
mechanical principies. Although man is capable of
rendering a wde variety of pattemed modcatons
00 bones it s suggested that hs primary strategy s
cutting, not tearing and pulling, and hamrnering, not
vslng down on obects. Aoimals menpulate bones
and carcasses by Iilerally teariog them down struc-
turally. This pattem s ene of gradual weakening and
with ts altendant sequenttal modtcaton of
weakened locations results in accumulated patterns
of both surcel and structural modtcaon of
bones. Man's slrategy could be cheracterzed as di-
rect. Using cutting tools he can echleve dternember-
menl or Hlleting directly wlthout firsl weakening the
joiols Ihrough an altrilional or destructive prelimi-
nary process. Animals cannot do Ihis, except
perhaps where Ihe prey or bone being manipuIated
lB much smaUer Ihan Ihe size of the jaws of Ihe active
animal. Uslng percussion lools, man can break bones
directly without Cirst weakening them through the
gradual destruction of their morphology prior to
brealdng. In a very general sense, the more an animal
acts on abone the more it is modified, and stroctural
destruction is an inevitable consequence of Ibis ac-
Iion. Man may dismember a careass, prepare the
meal for consumplion. and break the bones for mar-
fQW without necessarily deslroying bone material.
He may break it and dissipale the anatomical slmc-
ture but he does nol deslroy it, only renders it into
smaller pieces. As o normative slalement the con-
lrosl of animols as bone deslro}-'ers and meot "teor-
ers," with men as bone breokers ond meat cutters, is
probobly stotisticoJ/y occurole. This is whal I hava
referred to as an empirical generalization or an in-
ductive argument from enumeralion [sae L. R. Bin
ford 1978a). It is quite probable Ihal Ihis is an accu
rate generalization and could be restated to meel a\l
of Schiffer's (1976) crileria as a "law" lo be used in
archaeological inference. It would nol leed llS to a
secure knowledge of Ihe pasl, simply because there
isno condition of necessily shown to be determining
the cuslomary actions o{ man as generalized.
Man can produce bone lools by destroyiog se-
quenlially snd lo a paltero segments of bone
morphology-hence the ambiguily between animal-
gnawed bone and pseudolools. Man can butcher
animals using bludgeoning lechniques and techo
niques designed lo weaken Ihe anatomicai slruc-
ture of an animal gradually so Ihat the reslllts may
look very mllch lika the deslruclion caused by ani-
mals gnawing apart a cercess-c-ths s certainly tllus-
trated by Frtson's work. It ts elso proved by Frison's
replicative experimenta for there at leaet man did
mimic Ihe producte of scavenging antmals. Men's
flexible mechanical capabilities made poseble
through the Use of toola render the spectcaton of
any relationship of neceeety between man con-
cetved of as btclcgtcel species and the products of
his work almosi impossible. Man could under cero
tain condtttons replicete any mechanical conse-
quence of ether other animals' behevor or ecton by
physcal forees. We can imagine stuetons in which
tool-u5ing man might be forced lo such destmctive
chopping lechniques for disarticulating Ihe humerus
from Ihe scapula resulting in the deslmction of Ihe
proximal humerus (see Frison 1970; White 1953a;
and many others). We can imagine situations in
which 100I-using man might avan pick up abone
previousIy modified by an animal. for inslance, a
so-called humeros "flesher" (see Frisan 1974) and
use it as a flesher! Certainly replicative a;ll;perimeo-
ters have done Ihis or in sorne cases they have made
facsimiles of such bones and used them successfully.
Why not early man? The flexibility of our own re-
searchers to invent accommodalive models for all
the forms of deslruclion 1have shown to be regularly
produced by animals is sufficienl in itself to show
Ihat man could do these Ihings; there is no linkage of
necessity between man and a partieular foolprint in
Ihis case. Man seems lo be capable o, mechanieally
speaking. producing traces on bones that could
mimic any other "natural" footprint thus far known.
I would Iherefore suggest Ihat Ihe emprical
generalization offered, although not particularly use-
fui for interpreting the producls of human behavior,
can serve as a useful guide in the research designo
for inslance, if one observes a pattern of bone de
stroction and knows that destruction is the normal
consequences of animal behavior, one should view
as one's task Ihe disproof of the proposition that
animals were responsible for the observed pattern.
One approach might be to demonstrate the pattern to
exist independently of recognized evidence of ani
mal behavior Isuch as toolh scoring; pilting; chip-
ping back; production of chipped-end. longitudi.
nally fraclured diaphysis fragmenls; cylinders;
scooped out soft cancellous ends of long bones; and
billen off bone prolrusions such as Ihe grealer
trocbanter on the fmur]. Failure to follow this
stretegv is probably basic lo the errors made by Fri-
son (1970) as well as that of Stanford (1979a) et the
Selby and Dullon sttes end most of the workers at
Old Crow Flats [Bonnchsen 1979). One might sus-
pecl Ihal tbe reversa strategy might prove helpful
when a pattern of bona breakage or rnodficaton by
percusson s noted. Namelv.knowing that breekege
te a normal consequence of human bebevtor. ene
should vew as one's task Ihe disproof of the prcpos-
Iion that man was responslble.
This s a much more dfficult tesk. slnce we heve
alreedy suggested that man is cepeble of producng
modcetons that can mimic mosl of Ihe effects of
any natural process. In arder lo do lhis, we would
have to have exbaustive knowledge of alllhe possi.
ble modifications on bone of which "nature" was
capable, and then have performed sufficient
middle-range research to permil us to differentiate
aH these possible modifications from their nalural
For instance, it can be shown Ihat animals regu-
lady produce spiral fractures, Spiral fractures seem
to be the cbaracteristic resull of at least sorne types of
impact loading of bones, The very presence of spiral
fractures suggests that animals may also modify
bones in ways Ihat produced al least sorne formal
consequences similar to impact loading. For in-
slanee, 1 ha\le seen puoctures in bone shafts Ihat if
Ihey had been seen in isolation from the remainder
of Ihe assemblage, would compare favorably with
impacl nolches known lo resuH from Ibe use of
hammer in breaking marrow bones. Similarly, the
patterned modification 1 described in Chapter 3
as "mashed edges" bears many morphological
similarities to the impacl roseltes or "depressed frac
tures" produced by harnmers on thin-welJed bones.
We have a few studies of the breakage morphology
snd descriptions of the statistical dislribution of
fonos of scarring both from impact and from culling
on different segments of animal analomy. We do nol
yel have similar statistical descriptions {allhough
Hill [1980] has made a slarl in Ihis direclion) regard.
ing the placement and patlerns of associalion be
tween scarring and differing anatomical markers on
prey animals. These delailed descriptive malerals
are needed and may eventually serve lo permil ralher
specific identifications of agents. We do oot ha ve
4. Humen Morles of Bone Modificafion
this cepebtlity at Ihe present time because the ac-
tualistic middle-range studes have not been con.
This problem is perhaps well illustrated by the
current work of Richard Morlan (1981), mennoned
in Chapter 3. Morlan has correctly dismissed many
of the clairns that man was the causal agent of rnodt-
fied bones recovered from the Old Crow Flats local-
ity. However, as a consequence of his research he
further documented a pettem of hreakage previously
noted by Bonnichsen (1978, 1979), the patlerned
flaking and fracturing on segments of bones from
exttnct elephants. Morlan has correctly sought the
recognition of patterning manfest in ltems from
deposits remaining from the pasto This is fundamen-
tal and an essential part of archaeological research.
Having established that a paltem exists, we musl
face Ihe question of what it means. Morlan has rea
soned Ihat the bones of elephanls are loo large to
have been modified in this relatively robust manner
by carnivores, and that the patternng is too redun-
dant lo be the result of acddents of nature. There
fore, he has interpreled these bones as evidence for
man's participalion in Ihe events represenled in Ihe
admitled paleonlological deposits at Old Crow.
The reader wiJI undoubledly recognize this as a
c1assic argument from eliminalon. 11 might also be
recalled that for such an argumenl lo be correct the
assumption musl be met Ihat aH Ihe possible causes
have been lisled and aHbut one eliminated. To whaf
degree do you think we are capable of Iisting all Ihe
possible cause of broken and flaked elephant bones?
Given out currenl ignorance of taphonomic condi-
lioos surrounding the burial of elephant bones in
differing circumstances, it is impossible to meel Ihe
criteria for a valid argument from elimination. What
is needed is middle-range research aimed allearning
aboul elephanl bone taphonomy. This would in
dude actualislic sludies of dead elephant carcasses
and the behavior of other animals in the presence of
elephant bones (the effects of trampling, tossing
bones around as is recorded far elephants thero-
selves, and so on). Research mighl also look at
known paleontological assemblages where the pres-
ence of man can be Hlled out on historieal grounds to
see if such ilems occur io known lo be
unrelated to human action.<;. AII such studies must be
done by us. As staled earler 1know (lf llO olher sci-

ence concerned with explaining properties of Ihe ar-
chaeologcel record.
1 discussed in sorne delail Bonnichsen's (1973.
1979) use of ethnographic expenence among Ihe
Calling Lake Cree as the bass for identifymg a "hu-
man" pattem of marrow cracking. The weakness of
this approach is that one cannot argue for there being
a necessary connection between man considered as a
biological specles and cracking bones by the "rnid-
diaphysis smash technique" or butchertng bison by
"muscle stripping" etc. Equally cleer s the fael thet
men could and did break bones by other techniques.
Men uf the past may or may not have bulchered ani-
mals usng Frtson's technlques, yet man ts clearly
capable of doing so. as demonstrated by Frisan. AH
forms of arguments from analogy must attempt to
justify an assumption of necessi!y. If research suffi-
cienl to establish a necl?ssary conneclion had been
conducted, then an argumenl from analogy would
not be needed. This means that aH arguments from
analogy are subjeet to error. as are any other purely
indicative arguments. J suggest Ihal as long as such
arguments serve as Ihe basis fer a melhod of in
ference as was developed 8rollnd Frison and Bon
nichsen's work. then we operale with a sel of con-
ventions for "knowing" the past Ihat have not been
subjected to scienlific investigalion. This procedure
tesults most commonly in modern mylhs.
The traditional archaeological procedure was to
the archaeological record and seek to
documenl patternng (pattern-recognition studiesl.
Once patterns had been recognized. then one in
vented a post hoc accommodative argumenl, which
if true would account for the patterning observed. If
il appeared plausible and consistenl with whal was
oonsidcred to be our general knowledge of man and
nalure, Ihen Ihe post hoc argumenl would gradualJy
be accepted and become part of archaeological
theory. lo short. a convention would grow up such
that when ene observad the palteming origlnally iso-
lated by the inilial researcher one would interpret lt
[Le.. gtve meaning to it] by reetettng the original post
hoc argument. "Reliability testtng" normelly con-
sisted, under such ctrcumstences. of ponting out
how well Ihe Iacts lil the Iheory! Thts is uneccepta-
ble: we musl have middle-range research.
Implied in the metenals preaented thus far s a
principie that has not generally cheracterzed tradi-
tional epproaches. the implication Ihat a pattem that
may be recognteed by an ercheeologtst can Ire-
quently be expected to be the accumuiative conse-
quence of a variety of causal conditions. Presenting a
single scenario of the past as an accommodative aro
gument for Ihe palteming observed is Iikely to be
wrong, since the constituents of Ihe pattero are likely
to have arisen from different "causes' and lo be
combined in characteristic ways in terms of slill
additional determinant conditions. Gillen such ex
pectations, mddle-range research musl pravide Ihe
diagnostic 10015 for analysis of pallerning, thereby
providing information about a wide variety of past
condtions, ralher than the "one pattern, one in-
terpretation" form of most traditional archaeologcal
One might reasooably ask at this point that if we
cannol establish a peltern of bone modification un-
ambguously referable to man, why study the faunal
products of man aod seek greater understanding of
his highly variable behavor? The answer to this is
simply Ihallhe basic lask of anthropology-of which
archaeology IS a part-is lo seek an understanding of
man's variable cultural behavior, to explain cultural
smilarilies and dfferences.

:, --::-'r.2
-.! Ir
.... ,.,:l-lj:...';J.- ....
;;.0 .... <-- ....
, ....
Assemblage composition is a relatively new do-
man of reseerch. Wilhn lithic studtes Ihe rnost
common epprcech in the past has been cberacterteed
by Sackett (I968J as "the attempt lo equate erttfect
fossils. cultural stages. and geologtcal horizons as
narrowly as poseble into mecbanical systems whose
parte erttculated wtth almosl robonc precleon [Sac-
kett 1968:661." As Sackelt potnts out, the compara-
tve study of Ihe full range of asaociated lthc mete-
rtals. en assemblsge. has been an tnnoveuon largely
referable lo Francots Bordes and his colleagues in
France. 1suggest that Ihere was a secondary center of
develcpment in Africe stemmng largely from the
lnf'luanca of J. D. Clark. F. C. Howell, and Maxine
Kieindienst (see Vincenl 11978J for an nstghtful sur-
vey of Africanists' research).
The eerlest descdptton of a faunal assemblege
presentad as analomical par! frequences. of which I
arn awere. was the lruly seminal wnrk of Alfred Rus
(1937. 1943, 1958) and his collaborators Walter
Krause (19371 and Walter Kollau (Krause and Kollau
1943). lo these remarkable archaeological studtes the
fauna was described in detall and the relative fre-
quencies of analomical parts were summarized for
larga rendeer populations. These studtes have been
influenlial in a number of fields. for instano- in the
sludy o stte structure as acknowledged by Andr
Lerui-Dnurhan and Michel Brztllnn (1972:2461 in
Chapter 5
Pattems of
stemming from the
behavior of man
versus that of beast

5. Assrmblog!l Composilion: Pcttems of Associotion Strrnmil1g from !he BchIJvior uf Mon versus Thal uf Beost New World Regillnillgs-Man as the
their ptcneenng work at Pincevent. Despte the in-
noveve prtorty that the wcrks of Rust. Kreuse. and
Koltau certainly enjoy in the hstory of archaeology,
it was While's work [1952, 19538, 1953b, etc.] that
had an impact on the fieid so significan! that study-
ing Iaunal assamblagea in torms of anatomtcal perta
became a common approach. As will be described
later Dart's wnrk had a similar impact 00 old world
New World Beginnings--Man as the Agent
In North America lhe carlies! work 00 assemblage
composilion viewed frorn the perspeclive of anatom-
ical parts was done by Theodore White (1952, 19538,
1953b, 1954, 19551. The principies for interpreting
differential frequendes among anatomical parts
were slated in one of White's earliest p8pers:
Thll accompanying lable shows discrepancills which
Cllnnolbe Ilccounledfor by accident of It is
difficult to escape Ihe inferenee tha! lhe parls eilher
were not broughl into camp, as with the Choracie verta-
brae. or Ihal they were mutilaled beyond recogntion
while cutting up Ihe carcus, Le.. tha proximlll end of
the humerus White 1951:3371.
White considered variability in analomical part fre-
quencies lo be referable primarily to differences in
bulchering practices and secondarily lo Ihe dif
ferential abandonment of analomical parls in the
conlext of hunting logistics. He also mentioned the
selective use of bones as tools. As discussed earlier,
mosl of White's argumeols were posl hoc accommo-
dations to accouo! for Ihe patterns of anatomical var-
iability noled in his samples. His model ofprehisloric
butchering nduded an assumption of lhe regular
use of large chopping and hacking lools and a simple
assumption of site functions-villages versus kili
Afler While's nilial work, Ihere were several
comparalive sfudies of faunal assemblages. One of
the earliest oneS was concemed primarily wilh rec-
ognizing "ethnie in "butchlJrinR practices"
(Wood 1962). Following this early comparison
was a considerable hody of informalon compara-
lively summanzed in the report un the Boarding
Schoo! Bison Orive stte [Kehoe 19B7} end Bonfra
Shelter (Oibble and Lorrain 1968). Both reports pro-
vided revtews of ethnohistortcal and ethnographic
sources on bison procurement. descripons of the
sites including sumrnartes uf anatomcal part fre-
quenctes by p-ovenence unit. and fairly cornplete
comparisons of 'he authors' meterlels with pre-
viously published assembleges. In both cases most of
the assemblages available for comparison were from
the "vllages" studied by White. This meanl thal
Ihere was concern Cor the mirror-image argumenls
Ihal were implied by While's inferences regarding
how kill siles should look. As one mighl imagine,
there were sorne reasons for confirming and sorne for
rejecting White's arguments. Nevertheless. in both
cases there was a large body of variability, Ihe causes
of which were not apparent. There was sorne indulg.
ing in posl hoc argumentation. but in general there
was a kind of glossing over ,he variability; a "what
can you say?" atlilude was expressed, Although it
was nol discussed al the time, the demonstraled var-
iability signaled thal things were nol as simple as a
patterned conlrast belween kili sites on Ihe one hand
and villages on the other. Roughly conlemporary
with hese comparative studies on the American
plaios, which were largely using White's melhods.
Ihe first "exporls" oC White's procedures to olher
areas of the world appeared. By 1969 White's proce-
dures had been applied in the Near Easl and. as we
will see. in Africa.
Perhaps the most bizarre argument lo surface was
published by D. Perkins and P. Daly (1968) regarding
a "Neolithic" sile in Turkey [see also Daly 1969).
Oddly this is perhaps Ihe besl-knowll argu-
menl regardinEl fauna. These authors have been
called Ihe discovers oC Ihe "schlepp effec"" The
schlepp effecl has becn caller! a "Iaw" by Mike Schif
fer (1976:Z1j.lt s. however, not a law but a post hoc
accommooative argumen! thal seems lo have very
liule to recommend i1. It is not based on elhno-
graphic analogy; il is sheer accommodative [anlasy,
yel il has served as Ihe interprelative basis for a
number of argumenls aboul fauna.
wor\cingacl'aleomfian siles in lhe New
Worlrl have tha!. il nnly a few parls af a bisan
sll<!elon are fOllnrl. lhe sile is a more or less
pennsnent c:amp lo which the hunters brought back
meat from animals they harl klled and butcbered else-
where. lf on the contrary mas! of the btsons bolles are
present. the ste is prcbebly a "kili" arte lo which tbe
successful hu nters surnmoned therr families [Perkns
and Daly 1%8:104).
The rst thng wrong with this prcoostton is its
assumpton that all stes are either kills or rest-
dances. It also assumes thal decisions lo transpon
hones and proeess mear are always made in the same
wey[e.g.. we alwavs fillet meal, we always transport
only filleted meat). and Ihal Ihe decision lo transport
animal parts is not conditioned by either dislance to
becovered or the quantily of meal to be Iransporled.
A more detailed inlerprelative model was in-
vented by Perkins aod Daly lo accommodale Ihe spe-
cific facts thev observed in a Turkish site of Ihe
sixlyfifth century B.G. They observed that bones
from the feet and lower legs of cattll! were more
common in Ihe site Ihan were Ihe hones of the upper
When a Suberde hunling party killed a wild ox. t!ley
apparenfly butchered il on Ihe spol and used Ihe ani-
mal's own hide as a conll'liner for earryin: Ihe meat
home. They evidently slripped the forequarlers and
hindquarters of meat and Ihrew the leg bones awa)'
They appareolly lefl the feel Illtllched lo Ihe hide.
ptlrhaps because Ihe feet mll(ie convenient hl'lodlesfor
dragRingIhe meal-fllJednide. Perhaps Ihey also valued
Ihe feel; this parl of lhe animal contllins useful
Ilnd has been called "the hunler's sewing kit." We ha\'e
named the disparity between (he number of cllltle foo!
bones and leg bones Ihat resulted tram this treatment of
the prey the "schlepp effec\," after Ihe Germ..n verb
meaning lo drag /perkins and Dal)' 1968:104[.
Nowit ig very clear lhat lhe feet and lower Jegs yield
tillle if sny meat. A "principie" thal had also guided
Plains faunal interprelation from Ihe very beginning
was thal man aeted differently loward par!s thal
yielded low food relurns per unil weight. For in-
stanre, While slales. "Because oftheir greater weighl
the skull, scapula, pelvis, and melapodials of the
bison were not transported as far as the correspond-
ingelemcnls in Ihe deer and anlelope and were lefl
althe kili [While 1954;2561."
ff Parkins and Daly had trealed the Plains Iitera-
turemore fully Ihey would have realized lhat the two
basic propostttons were in contct in their case: The
number of bnnes was relatively small bUI tboee
bones present ware from low-util ty parts-cyet they
belteved they were excavatng a resldental stte.
They gnored the more baste propositlon regardirtg
transport , nemely that one preterenally moves lo
pomts of consumption perts with high meat yield
per unit welghf and thev Invented a "just-so story"
to eccommodate the Iacts and still preserve Iheir be-
Iiefs that (a) these were Neohthc hunters and lb) il
was a residential sue or village. This Is mytn moking
of !he firs! order.
The amazing thing is Ihal people adopted the
schlepp effecl as a melhodological principIe. For in-
stanca, the small cave of Palegawra (approximately 5
m deep and 6 m acrossJ. sorne 70 m aboye lhe valley
Ooor, was nlerpreled as a "residential site" by vir!ue
of Ihe implicalions of Ihe schlepp efect.
The relalively high proportion of numoers of individuo
als which would have been counted on the basis of
mandibles and distal bones oI !he limbs. as compared
with maxillae and more proximal bones of Ihe limbs, is
apparent the disproportion of Ihese bones we altrib
ute to "sehlepping" . !he favorite bUlcheting lech-
niques with a dead ungulale seems lo have been to skin
it where iI IeU,leaving the feet and mandibles (iess often
Ihll maxillae also) in the s\cio.The meal was then cul off
Ihe carcass and plled on Ihe spread skin, after which Ihe
feet and j8W provided hanrlles for sometimes dragging.
and sometimes carrying Ihe load hack to the cave. This
ellplanation fjts the faets as we find them al Pategawra
[Turnbull and Reed 1974:136-137).
Again. lo imagine Ihe past in this manner is sheer
rnyth. The piclure of the ancient Zarzian hunters
dragging and pulling a skin full oC meat 70 ro up to
Ihair "residence" is probably as realistic as imaElin-
ing the meal being transported lo Ihe cave by andent
aslronauls. The Cactsal Palegawra had been "fit into"
Ihe assumed site Iypology of residences and kili
sites. The hequencies given for Palegawra. although
nol well described, appear much more comparable to
Ihose reported by Raymund Wood (1968) from Vista
Shelter. In a very importanl paper, Wood explored
the idea of orElanizationally more complex
subsistence-seltlemenl sysfems.
Wood argued Ihal the faunal remains oC deer and
bison found in Vista Shelter had to be viewed as
5. Assemblage ComposHion: Pnnerns of Associuljon gtemnung Irom the Behnvlor uf MUl> versus Tha! of Becst New Wor!d BellJrlOlngs-Man as rhe Age/11 187
derivad from Ihe acttvttes of hunting parties
ortginating in Ierge permenent agricultura! settle-
mente miles away from the nunttng grounds in
which the shelter was Ioceted. In addilion to
suggesng that there could welI be other Iypes of
sites with other ohsrectenstc faunal assembleges.
Wood's study pulnted to facts Ihat had frequently
been ignorad in eerher studies: "All deer and btson
humeri and redil are fragmentary, Ihe sha having
beea shattered by heevy blows, leevlng onty the ar-
ticular ends naarly ntect. ThA stubs ot the shafts
aften show pitting resulting {romthese blows IWood
1968:1751," In tenns oC the descriptions presented
here, the pits mentioned by Wood were most cer-
tainly impact depressions. Wood <lIso noted bulcher-
ing marks snd providel'i sorne descriptions oC these
marks. He waS arguing for a hunting camp; al least
one more type of site WBI'i acknowJedged. Although
Wood's work represents a majar slep forward, his
line of im'estigation was not developed further lor
sorne years.
In Africsn sludies, Brian FagilIl (1967: Fagan el
ojo 1969) reports the assemblagB composilion for
both wild tlud domestic animal6 from a numher of
Iron Age sites in Zambia. Little interpretation is oC
fered and differences noted belween sites are gener-
aUy referred lo differeneas in butehering practice as
manifeststions of minar cultural differenees. Of pro-
phetic importance however, were two Ihings: (o) lhe
recognition that Ihere were major diCferences in the
relativa frequencies of anatomical parl!! at lha same
site among spedes of differing sizes {sce Fagan
1967:83}, snd (b) the recognition oC the problem o
preservation: "Anolher unknown factor is that af dU
ferenlial preservation. The relative resistance of dif-
ferenl body parts lo deslruetion by natural means is
unknown, and it mav be that our distributions are a
refleetion oC rather than habit [Fagan
1967:82)." This Iype of skepticism, in the context of
the clear inability of White's modals to offer
guidelines for intcrpreting tha wide variety of pat-
tems being revealed as more descriptiva studies ap-
peared, surfaced in America as well. For in-
stanee. Pillaert {1969j analyzed Ihe whitetail door
bOD from lhe Mi1Iville site in Wisconsin and con-
cluded, "the percentages do not produce a paltern
lhat would beconducive lo determining Ihe bulcher-
ing process [1969:1011."
Perhaps the most crucial objection Ihat bagan to
be voiced regerdtng White's methods had rererence
ro hs assumption about preservalion. While had ac-
knowledged thet the counts of bone perta might be
suhject te verietton as a result of dterences in pres-
arvaton. but suggeeted that each part hed an equal
chance of belng preserved. so Ih{' MNI ratos would
be an eccurete reectton of the parts origfnally pre-
senl on the sile (While 1952:337)lohn Culday [n.d,
and personal oornmunicationJ voiced the pesstmtstfc
vtew thet all the variabilily in anatomical part fre-
queodes was related to Ihe differenlial destruetion
of anatomical parts by "village dogs."
This was essentially the resealeh environment
during which imporlanl work by loe Ben Wheal was
initlated. Wheat was finding the unmistakable re-
maios of bison thal had boon driven inta a small
arroyo. Unlike lhe site dug by Kehoe, 0lsen-
Chuhbuck appeared lo represent a single episode.
Thc resolution of Ihe was high, in Ihat only a
few sequentially related events were representad,
uncomplicated by Ihe buildup ofremains from many
events, almost certainly the case at Ihe Boarding
School site (Kehoe 1967). Wheat's excavation
strategy was sufficient to permil the recognHion ol
sorne very important associalions in (he deposit.
As we e"r..avaled 'he bone depo&il we finf uncovered
Ihe upper layer G<lnlainin.lllbc singlc and ltl1icu-
lated segmenls of skelelon. It was soon apparent tbat
these bones were Ihe end resull of a stolldordized
Pofeo-f"dion butcnering procedUffl. We carneto recog-
nile certain "butchering units" suchas forelegs, pelvic
girdles, hind legs. spnal columns end skulls. Unil5 ti
the same kind weUl usually found together in groups
numbering fromtwo or Ihree to as many as 27.Similar
unfs alsoformed dbtincl vertical sequences.. . Where
thll order t)f depostion ws.s cleat, the bones al Ibe bot-
tom of e8chpile wereforelegunils. Abovethese bones
werethoseof pelvicgirdleunits. Sometimesoneor bolh
hind legs wereallached lo Ihe pelvicgirdle, but bylllld
large the hind-Ieg units lay sepllralaly amongOf abova
Ihe pl::llvic unls. The nexl level WIlS usually composed
of spinal-columnunits ... litthe lop of neerlyeverypila
were skulls. The jawbones had bllenremovedfroID most
of Ihem. bul sornesllll relalned R fewof lhe peckverte-
brae IWheat 1967;149-501
Wheal was recognizing an associated pattern.
both in the character of bones remaining articulaled
and in tha anatomical associaons belween the rec-
ognized units and their vertical dsposttton within
the narrow arroyo. Wheat's nterpretation of this pat-
tem was justified by cnng generalizations sdvenced
after an extenetve survey of the ethnohistorieal
sources descrtbtng Plains butchering procedures.
(Sea Wheat (19721 fur a review of Ihis Itterature.) In
shcrt. wheet's tnference as lo rhe behevorel mean-
ing of his ndlngs and in turn hs reccnstrucon of
the butcherng prccedure used over 8501) yeara ago
at Olsen-Chubbuck was a classlc argument from
elhnographic Bnalogy: "jt reasonable to as-
sume thal the Paleo-Indians foHowed the same ini-
tial steps in bulchering that Ihe Plains Indians did in
recent times IWheal 1967:501."
Que uf the steps in the prot.;edUHJ as describecl
elhnohislorically was the removal of Ihe "blanket of
Direclly under the of Ihe backwas a laverof lender
mea\, Ihe "blankel of flash"; when Ihis was stripped
away. the bison's forelegs and shoulder bladescould be
cut free eposng lhe highly priled "hump" meato Ihe
rb cage and Ihe bodycavi!y IWheat 1967:50J.
This model of Ihe pasl was strikingly different
from the model of butchering presented by While,
whete tbere was an assumption of regular use of
chopper!> aud deevers with the consequent destrue-
Iion of analomical parls during butchering oper-
alions. No 5uch picture was presented by Wheal's
work. One obtained the impression thal the butcher-
ing was primarily done by eulting Ihrough the joints,
and dJOpping aud smashing were nol so abviously
The nexl most important development in Plains
research was lhe work of George Frison (1970}, pre-
viously cited. His nteresl in butchering praclicell
waS slated from the very onsel of his analysis of Ihe
Glenrod. Buffalo lump, but his approach was dif-
ferent from that of ether White and his "followers"
such as nihble and Lorrain (1968) Bnd Kehoe [1967),
or loe Ben Wheal (1967) who was working wilh
ethnohistory and units of articulBtion. Frison
(1970:8-1O} msde basicaJly tWD argumenls: (at that
alllhe marks or modifications on the bones al Ihe
Glenrock jump had been produced by Ihe "tools"
remaining at the sile. and (b) Ihat tools were recog-
nized by virtue o a patterned form of breakage, a
petterned Iorm of structural modification, and/or a
patterned Iorm of surctal scerring. Ths leads di-
rectly to tha correlafAd argument cf there hF!ing many
bone Iccis al Glenrock.
1 have alreedy Ireated Frison'a work al Glenrock
in sorne detetl.It Is cleer lo me that the btson bones at
Glenrock had been ravaged by predatcr-scavengers.
Many of the modifications and much of the dstruc-
tion manifesl on the bones wes a censequence oi
nonhuman animal behavor. Frison, however. oper-
eung on Ihe assumption that aU modifica\ions were
human, invented a post hoc accornmodative
argument-muscle stripping executed largely wilh
bone too1s. Frison's knowledge of bolh 1001
mechanics and animal anatomy !Ierved lo ensure thal
his model was in fact fcosible. When he butchorcd
bison in a replicalive experimenl. he col.lld produce
many of the modifications observed on bones al
Clemod. This demonstration was convincing to
bolh Frison Rnd many olhers. Among a group of
largely Plains and early man enthusiasts the replicat-
ive experimenl hecame an epistemologicsl clevice. A
c1assic statemenl heard al meetings in reply lo some-
one's sUBgeslon as to how Ihings were done in the
pasl was Ihe quip. "How msny limes have you done
that?" If the answer was "never" Ihen a knowing
glance drculated among lhe in-group and Ihe person
snd his jdea were dismissed as being not worthy of
further discusslon (see the exchauge between St<lJl-
ford and lohnson in Davis end Wilson 11978:3101l.
This "going native" approach is ilIustraled by Ihe
following slatemenl from Frisan: "hunling can never
be fully underslood in on onthropological.tense until
the investigator has had lo fjl his empty stomach by
means of his own expertise and ability as a hunler
/Frison 1978:3641."
Tha American public has been exposed lo pholo-
graph!l of Dennis Slanford using sn atlall {Slanford
1979bj, butchering en elephant {Park 197B), and eal-
ing bison meat "EsJdmo slyle" (Odyssey film, Seek-
ing Ihe Firs! Americans, 1980)-presumllbly under
the assumption thal al1 lhese things establish his
creclentials a!l an experl on hunling and gathering.
We may also assume !hat all these same achieve
ments provide the basis Cor "understanding" the
"firsl Americans." Why should tha way 1 do some-
thing today in Ihe context of a kind of "boy scouf'
romanlicism or even while "lrying lo fill my belly"
5. Assemhlcge Cumposilion: Pnerns uf Associofion Stemming rom tho Rf'holior ur Moa versus Tho! uf Baos! AfrJcan Problems cnd Assemblage Composition 189
ensure that the people of the past did thtngs in simi-
lar ways? To argue thet they did ts lo argue from
egographic enciogy. 1
For nstance. George Frisen dtscussed his recon-
struoton o butchering, which was based 00 his irn-
egtnatton and replicative experiment. relativa to Ihe
reconstructons thet Ice Ben Wheat had oered for
Olsen-Chubbuck based 00 ethnogrephc analogy.
Reasoning from egographc enatogv. Frison
(1970:4Q) discussed wheat's "blanket of Ilesh":
It is diffic;ult lo understand wbet thte so cetled "blanket
of flesh" migh! refer lo unless iI is the thin iayer of esh
thal edheres lo Ihe hide during skinning. Actual butch
eringof buffalof8ib; lo revealanylhing Ihal salisfiedIh,
description. It was nol possibllltofind any waylo "worl.
Up under" any blwhl ofllesh lo removeIhe fronl q\lllf-
len jFrison 1970:401.
Ido not doubt that Frisan could nol find the blanket
of flesh but it was not because such a blanket did not
exist; it was simply because Frisan did not know
how to butcher a bison in the appropriate manner.
See, for instance, my iIIustratians of the "blanket of
flesh" removed by the Nunamiut Eskimo when
butchering far drying (Binford 1978b:95, Figure 3,
2). Cerlainly a bison could be butchered in an analo-
gous fashion. If we take knowledge of our "wlture"
as the basis for knowledge in general. or as Ihe stan-
dard for judging what is plausible or possible. then
we commit the worse kind of ethnocentric error.
Unfortunately, the approach ol Frisan has tended
lo somewhat dominale faunal studies in the
iI am in no wsy implying that practical experience,
familiaritywilhone's sub;ect matler, and raplicativeaxper
imenls arEl nol of value lo a researcher in an educational
senS6. ThllY are very importanl IIndshould be slronglyen
cou[aged. I am only suggeslingthal ji would be11veryrare
Ihing iodead if any of us could become "completely edu
celad" bysuch self-teachingtactics. I wouldbe verysl.epli-
cel of the claim that tlllY one experl was knowledgeeble
about Ihe full range of behavior Ihal ffillnkind in general
had perfeclp.d and execuled in diffp.ring drcumslances,
many of which are cerlaln lo be unapprecialed by Ihe
single individual WOl"king from even a very complele
knowledge uf his J8rlkula[ experiences.
Americanisl literatura in recent years. Assemblage
vartabtlty Ireated in terrns of analomical part freo
quenctes has not been sludied recently on the Pleins
even at the pattem-recognttton lovel of rasearch.
Those who have tended to be unimpressed by Frison
have quietly sided wlth loe Ben Whp.al. For Instance.
a student at a Platns universlty only test year SIl-
nauneed te me that "we don'l study anatomical part
frequencies or breakagu. we study the artculated
segments of the anatomy as a clue lo butchering
practces.:' Anolher studenl at the sarna university
confided that "we think loe Ben Wheal was rigbt
about Plains butchering."
l must admit thal 1loo "Ihink loe Ben Wheat was
right" and lhat muscle stripping probably new'r was
performed on the Plains, at least not in Ihe manner
invented by Frisan. In Ihis case ethnographic anal-
ogy is likely lo be much more relaible than egog-
raphic analogy. The sad faet is, however, thal
anatomical part frequencies can reveal much more
Ihan just bulchering practices. In facl, Ihey probably
reveal very Httle directly about butchering practices
per se. since cul marks, palterns of articulation, and
patlerns nf context association are all apt to yield
more direct information about bulchering Ihan are
relative frequencies of analomical parts. White's
original assumption thal the absence of certain parls
was referable lo Iheir destruction during butchering
now no longer appears plausbile as a general condi-
ton. lt became clear thal Ihere are maoy pattems; in
fact, almost every new assemblage described and
compared-for example. Woud (1962), Dibble and
Lorrain (1968), and Kehoe (1967)-had unique prop-
erties. This wide range of variability was hard lo ae
commodllte to the simple model of seltlemenl Ihal
White and other early students offaunal assemblages
lended to imagine.
The major methodological implicalion of the
sludy of anatomica! part frequendes for New World
studies was simply Ihat one could lell the differenee
between kili and residenlial sites. lo fact there was
little difficulty in accomplishing Ihis lask on the
American pJains when bison was Ihe major animal
involved. 11 roa)" be that this banal characterislie of
the melhodological implicaliolls of studyinR faunal
assemblages contribuled lo the lack (Jf inleresl Ihal
seems lo have characlerized the work of the 19705
African Problems and Assemblage
Only several years afler the works of White had
turned the attenlion of sorne New World ar-
chaeologsts toward Ihe potenttal of anatomcal part
frequenctes as sources of informatioo. Raymond Dart
(1957) published a very important study in which he
deseribed properties of the Makapansgat fauna from
the perspective of enatomcal part Irequencies. Dart
had previously argued from patterns uf breakage,
partcularly of baboon skulls. thet Auslrolapithecus
was a hunter, and that Ihe faunal remains in the de-
posits from which the early fossils had been obtaioed
were "kitchen.middens" (Dar! 1949:2). The 1957 re-
port on his sludies of the fauna frorn Makapansgal
included a tabulation of relative frequencies of
anatomical parls. The variability in parl frequencies
noled by Dar! was interpreled as deriving from Ihe
selective removal of bones from kills far use as tools
by Austrolopilhecus.
The disappearance of lails was probablydve to their use
as signals and whips in hunling oulside Ihe cavern.
Caudal and olher vertebrae may also have disappeared
becaus.. of the polentiat value of !heir bodies as pro-
jp.cliles and Iheir pTor.esses liS ..vers and poinls ... fem-
oTa IInd libiae wouldbe lhe heavieslclubs lo use oulsidp.
Ihe cavp.rn. thal is prnbably why Ihese bones are Ihp.
leasl r.:ommon. Humeri are Ihe commonesl of Ihe long
bones: probtlbly because Ihey would be the mosl f.On-
venienldubs for Ihe women-folk and children lo use al
home IDart 1957:851.
Although Ihese inlerprelalive arguments appear
mylh-like today, Ihey were central in prompting a
great deal of imporlanl research. As pointed out ear
lier, it was largely Ihe daims of Raymond Darl re-
garding Ihe nature of the behavior of the early
hominicls Ihal demanded Ihal behavioral queslions
be answl;lred using archaeological remains. Actualis-
tic resear<.h was prompted by alternative arguments
as Washburn's (1957) Ihat Ihl;l earl" hominids
had 1I0t accumulall;ln Ihe neposils bul WPTl;l
present by virtue of having been prey lo wnalever
prenalor-sr.avpnger was responsible fm Ihe deposits.
Argumenls over lhrJ rtgent respomible fm lhe depo-
sts, the criteria for recognizing tools, and what
would be acceptable evdence of eerly hominid
hunling were all questcne thet stemmed from the
controversy surrounding Dert's clatms. This s a very
differenl intel/ectual envirooment from one that
seeks lo answer questions such as these: Was the
Mousterlen the cultural encestor of Ihe Aurignacian?
Out of whlch "tradltlon." Clactonian or Acheulian,
did the Quina Mousterian develop? lnstead of tradi-
tional historlcal questtons. the researchers of early
man were faced with controversy regarding the
charaoter of the early hominids' behavior, and Ihey
had to seek answers through means that were out-
side the dornain of traditional archaeological re
search procedure. 1 can remember leaching das!les
on Lower Pleistocene materials in the early 1960s
and pointing out that although 1thought Dart's post
hoc arguments, such as the one just quoted, were
sheer fantasy, the patterning he demonstrated at
Makapansgat was ioteresting and essentially unan
swered. That is, we did nol have reliable methods for
interpreting such patterning. Why were the anatom-
ical part frequencies al Makapansgat so different
from the frequencies in a living animal? Here was a
classic pattern-recognition sludy. Whal we needed
was middle-range research to investigate the causes
of such palterning. lf we isolated the causes we
mighl have !he basis for a powerful set of archaeolog-
ical methods. The Makapansgat fauna seemed to me
lo he a critically r.hallenging phenomenon. For in-
stance, I had bl;lcome accustomed lo archaeologists
"discovering" new paneros in thearchaeologcal rec-
ord but in most cases 5uch palterns were cited as
distinclive of Ihe people who produced them. There
was no reason to view Ihem as "slrange" or unex-
pecled since, ater all. cllllures varied-we all know
this. On the other hand, the daims of Dart were dif-
(eren\. He was comparing the frequendes of anatom-
ir:al parls observed in the archaeological record to
"expecled" ([equendes derived from a knowledge of
lhe anatomy of animals. When the observed differed
{rom Ihe expected he demanded an explanation-
what behavior, events, or causal condifions could
have "happened in hslory" lo produce Ihe dif
ferences belween lhe expocled conditioll and Ihe ob-
served facts? He was ble to make a uniformilarian
assumplion mgllrding the analomy of animals. He
5. A55emblage Composton: Pottern5 01 A550cilllioCl Sternminglrom he Behal'ior of Mnll venus That 01Beast

African Problems (lnd A5semhloge Cnmposttfnn 191
was eble to eay wilh sorne confidence what the fre-
quences o enetomlcal parts had been all tbose
thousands of years before. He wes therefore eble to
esk a very specific queston about sorne events tha!
we knew had to have taken place. namelv those
events that resultad in Ihe modificaton of the
anatomical part trequenctes away from the known
prior conduon.
It is true tbat essentially the serna etructural sttua-
tlon cheractertzed the earlter arguments by Tbeodore
whte. But there was a derence. White phrased the
comparison as a bass lar recognzing cultural dif-
ferences. It was ViAWed as a way oC monitoring
"ethnic history," and specific explanations were not
demanded as to the bahaviors responsibla for the dil-
ferences. It is true that a number of post hoc argu-
menls were offftred. bul one 1I0t Ihe impresson Ihat
the truth of 5uch arguments was not as important aS
the demonstration of cuhural differences, as was
suggested in the comparative studies by Wood
(19112) and Dibble and Lorrain (1968). This was 01
course Ihe procedure of traditionaJ archaeolagy;
patterning varied hecause it reflected cultural dif-
ferences. Tha culture varied as a fundion of the ac-
ddenls of history, The relationships hetween be-
havior snd the arehaeological record were not a
real1y serious concern. Darl's argumenl demonded o
behaviorol evoluotjon. He was saying that the
anatomical parts varied lrom the expecled frequen-
cies because of spedfic behaviors carred out by spe
cinc agents, and that these behaviors were distinc
tive to Ihose agenh, the early hominids. Controversy
here had to be wilh behavioral models, not vague
"historicaI" models. as to which archaeologically
defined unil influenced another or where certain
traits were invented and where they diffused. In Ihe
Dart case ODe had lo argua over (a) how hominids
behaved and (b) what archaeological feets were un
ambiguous indicetors of slleh behaviors. This was
Ihe intellectual c!irnate of New ArchaeoJogy, 8nd
hardly anyone recognized it!
Neverlhelp.ss. reSf!arch shifted rapidly to aclUalis-
tc studies. There was a major sel of research pushes
with contemporary primate studies, and Washburn
and DeVore were inslrumenlal in slarling what miht
be called "ntodern" hunler-gatherer research con-
sciously considered of relevance lo underslanding
something aboullhe chanH.:ler of early hominid life-
ways. (Sea the introduclory statament to Mon the
Hunter [Lea and Devore 10081)
Qne of the eerltest reseercb "spin-offs" from
Dert's cJaims Iocused on the behavior of the African
porcupne. Based en analogy with Dert's cletms for
bone tools at Mekepansgat, similar clams were
made for ttems from the Hopeeld site: Ronald
Snger (1956) reported thet those "tools'' could be
understood as products of gnawing by the Africlln
porcupine. Dart (1958) ccuntered by rst illustratlng
ethnographieaily documented bone tcols. showing
Ihat man did use bone for tools, and then iJlustraling
alleged bone tools from a site excavated by R. .
Mason called Kalkbank (R. J. MaSOIl et 01. 1956).
This site was dated by carbon14 and by Ihe style of
the few stone tools found there to around 15,000
yeal5 ago. Dart and Kilching daimed there to be a
very large number of bone lools present, arguing lhal
the loealion was a living site in an area where tbele
was Iinle available stone, therefore the occupants
had used bone as a raw material for 1001produclion.
They recognlzed !hat the Africen poreupne has been
active in modifying the bones at Kalkbank (over 25%
of the total bones were considered pore:upine
gnawed) but they insisted that this gnawing took
place after the bOlles had been brokcn or otherwise
modifil"d by mano In many cases modification prior
to porcupine alteration d()es seem to have baen Ihe
case. The question remainad, who broke or modified
the bones befure lhe poreupinc did its Of
course Dart and Kitching argued that man did. They
illuslrated spiral fractures and longitudinally splil
banes, which may well have been broken by man or
by predatur-scavengers for marrow. Darl ano his
companion assumed that man broke {he banes in he
observed manner in order lo obtain tools (see Dar{
1958). In addilion, Ihey interpreled the observed dil-
ferenlial freI.Juency of anotomic.al parts as
from selectivity on the parl of man in relurning parls
oE polential use as tools lo hs living site. Later Hen-
dey and Singer(1965j reported on a porcupine lair as
weH as the conlenls of !ln adjaeent sile occllpied
periodically by mano lt was noted Ibat there was a
major difference in the degree of fragmentalion
characleristic of lhe banes (mm lhe human sile ver-
sus Ihuse flum the procupinc lair. The olhpr point of
conlras! was in Ihf! frelluency of idf!nlifiablp. por-
cupille gnawing-.3% in IhfJ human site, bul an iro-
pressve 60.0% of (he bcnes in the prccuplne ette.
Ihis tnterest in the African porcuplne as an egent of
both bone modification and bone accumulation con-
linues today (see Brain 1980). This, however. is get-
ting somewhat ahead of the story.
AHhough Ihe pcrcuprne conecversv led to the
study of contemporary behavoral dynamics of por-
cupines. the methodology that I suggested as the
only one reetly appropriate to beginning middle-
range reseerch. the studies of more drect relevence
loessemblege composttton vtewed Irom 'he perspec-
tive of anatomical part frequendes. were carried out
by C. K. Brain. Brain's research directIy addressed
importanl questions: What caused Ihe differen-
Halfrequencies of anatomical parts observed by Dart
(Brain 1967b. 1969, 1976b)? What agent was respon-
sible for the accumulafion of Ihe deposHs al sorne of
lhe sltes yleldillg early homlnld bones (Brain 1968,
1970, J976a, 1980F
Brain carrled out actualistic studies oI modern
leopard behavior and observed a number of distinc-
tlve eharaClerlsllu: tllat (;UnVillCed him thal, al leasl
al the site of Swartkrans, leopards. not Ihe early
hominids, had been Ihe agents responsible ror ae;.
cUffiulating Ihe faunal deposits. The study that did
[llore tu l:hange Ihe direetioll of llluch Aldean re-
iearch than almost any olher was his e!hnoar-
chaeological study of the faunal assemblages occur-
ring at contemporary Hollenlot vilJages {Brain
1907b, 1969. 197Gb). In this actualistic setting Oraln
was able lo oblain a number of contra/s:
1. the Hottentot were goal herders; a11 animals
killed for food were initially presenl on the site as
romplete animals.
2. Once Ihe Hottentol were finished with the
bones Ihey were discarded aboul Ihe camp and free
roaming dogs scavenged the garbage.
3. Collections from Ihese sites yielded popula-
lions oE bunes Ihat ""ere exlremeJy variable io lhe
relative Irequencies of anatomical parts represented.
Sra;o rcm>aned Ihal since Ihe completo skololon
had been presenl to begin wilh, Ihe differential fre-
quency could nol be due lo selective inlroduc1ion lo
Ihe site, as [)art's ar.:umcnls would demando Inslead
Ihny had lo be duc lo r1ifforonlial dolotion fmm tho
populalion, Arain nofp.rl Ihal lhe bonp.s Ihal were
underrepresenled were soft aod characteri7.erl by
cancellous ttssue. He argued that the sequence of
eptphysal unten would differenliaIly eect the sus-
ceptibility of a parl fo destrucnon gtven a variable
eged populaton. He then prcposed that the dif-
ferenttel destmcton of anetomtcel parts as 8 fune-
non of ther tntnnstc strengtb by me hee roamtng
doge of the Hottentot camp was responsfble for the
differential enatomtcal part Irequenctes. Bratn next
compared his Hottentot goet samples to the relatve
frequendes reportad by Dart fram Makapansgat,
demonslralng an impresetve simlarity:
The evidence of tha Kuisab Rivar Soal bonalJ slronsly
SU8Se! thl1t the diproportlons which Dart encoun-
leJed do, In fael. nol require any spedal explanation.
Artificial seleclion of certain skaletaJ pal1s netld no! be
postulaled. Ir for instanca, antelope were hunled as they
carnelo drlnk at a waterhole in the entrance of Ihe cave
and Ihen con8umad by aUSlnllopithednes l:md
5cavenging carnivores. a considerable bone accumula-
tion eould have been built up in Ihe lower parls of lhe
cavetrl. The bonfl'l preserved would haya btlen those
oosl abie tu SU1Vi\'8 Ihe deshudlve lrelltm8nt tu which
they had been subjecled 11969:221.
This is a classic piece of middle.range research.
8rain was able lo gain inform&lional coolrol over Iho
dynamic candtions in terms of which fhe bone as-
semblage was modified. No selective introduction of
parts eharacterized the situation. Yet theresult was 8
highly variable set of differentlal froquencios for the
anatomical parls remaining on the siles. This had lo
derive from a series of deletions from the as-
semblage. Brain then went on to seek an understand
ing oE !he factoi"S conditioning the removal PfQ_
One could almost datect a great sign of relief com-
ing lrom aarly man research teams in Africa. Isaac
(1971) considorod Ihe problem of Darl's ciaims for
Makapansgat solved and in no need 01' further study.
Brain's findings were gladly inemporated inla Ihe
discussion of 01duvai Gorge (M. D. Leakey
1971:275-281). Mary Leakey begins her discussion
of fauna by poinling out Ihat Ihe disproporlionale
frequencies of parls of Ihe skeleton had long becn a
"puzzling feature" Ihal had never "been satisfactor.
Uy explained in Ihe pas\.o' She then summari:zp'l
Brain's Hullentol work and concludes, "There is
thus a combination of human and scavenger activity
5, Assemblnge Compositioo; Potterns of assoctction Stemming jrum rhe Behavior Df Mon versus That af Beas1

Afrkun Probiems onrl t\ssemb/o8" Compostion

similar fa conditions al eerlv living sttes where
hyeenas and ether ecevengers are known lo heve
gnawed the bones discarded by the hominid popula-
ticn 1M.D. Leekey 1971:2771."
Thls concluson incorporales en mteresttng as-
sumption ebout Olduvai Corge. namely that the de-
struction of anatomcal parls al her "living sites" by
hyenaa and others occurred after the homtntd popu-
laton had dscarded the bones. Presumably Ihis
would be eer the site was does
Mary Leekey know this" A tabla of compartson is
presentad illustrating a combined sample froro Dld-
uval Gorge, Iha MlIkapansgat bovids, and the Hotten-
tol goal sample. lt is concluded (hat they are all very
similar, and although minor contrasts are eviden!
Leakey is comfortable with the conclusion that aH
are the result of essentially the same causal condi-
tions: human utili:r.ation, with subsequent scBveng-
ing by prndator--scavengers. L.eakey does not con-
sider the added assumption that the animals present
were originally complete skeletons, or that Brain's
work iIIustrates no faunal characteristics referable to
human utilizalion in 50 faeas differential anatomical
part frequencies are wncemed, How can one distin
guish an assemblage ofbones ravaged exclusively by
animals from one where man had played a role? No
one had addressed this question. Man's role was
being assumed in aU the interpretations, both of
Makapansgat by Brain (1969), and Olduvai Gorge by
Mary Leakey (1971). We had no procedure for recog-
nizing the consequences al man's behavior, particu-
larly if one discounted spiral fractures, as was done
by Mary Leakey (1971:277-278). Was man bringing
whole animals back to his living sites alter success-
fui hunts? Ol course this is what one would have to
imagine if Mary Leakey's arguments were accepted.
Al least Brain had the animals coming to
Makapansg8t and being killed on the site; no such
convenienl reasons seemed available to account for
complete animals on the "living siles" of the early
hominids al Olduvai Gorge! This implicalion of
adopling Brain's argument seems lo have been con-
veniently overlooked.
The next majordevelopment in the ongoing sludy
of assemblage variabilily from an African pt'irspec-
tive carne from the work of Calhy Read-Martin and
Dwighl Read (1975). They Ihar sorne insighl
eould be Rained on the Makapansgal siluation by
comparing the profile of anetomfcel part frequences
from the North German sttes originally reponed by
Rust's collaboretcrs Krause (1937) and Krause and
Kollau (1943) with the Makapensgat prole. Preser-
vatlon was reporled lo be excellenl at the German
sitas of Slellmoor and Meiendorf. 1 wculd agree.
There appears to be no suggestlon thal anv post-
depositiunal destructive process was al work In the
German stes. The faunal assemblages seem to repre-
sent unmodfied primary depositicnal assemblages.
The character of the German fauna was Interpreted
as represenling primary butchering near the km lo-
cation as welJ as Ihe foad consumed in support o the
hunling group whiJe they engaged in Ihe hunling-
butchering. The variable anatomicel part frequencies
on the German siles were interpreled in a way simi-
larlo the models argued in the American literature:
Differential frequencies of anatomical parts reflected
differential Iransport or abandonmenl of anatomical
units in the contexl of hunting logistics. Comparisoo
of the German sites with Makapansgal revealed thal
in the latter site scapulae, pelves, sorne vertebrae,
and ribs were rela1ively underrepresenled. On the
othar hand, heads and sorne parts of the front limbs
were relatively more abundant. II also appeared thal
Iimb sections were introduced lo !he MakapansglJ!
JocaUon after having been disarticulated froro Ihe
pelvis or Ihe scapula. These comparisons were made
despite Brain's demonslralion of a slrong set of cor-
respondences between the profile of parl frequencies
at Makapansgat and Ihe ravaged assemblage froro
the Hollenlol villages. There was strong support for
the recognition ol the Makapansgat assemblages as
having suffered considerable destructive allrilion,
and no such evidence of r1eslruction was characteris
tic of Ihe assemblages from Ihe Gormao siles. Thal
there was a difference belween Mak.apansgal and Ihe
Cerman sites wes dear. That Ihe difrerences were
referable to the original composition of the as-
semblages was nol clear. The ravaged charader of
Ihe Makapansgat assemblage had becn convincingly
demonslraled by Brain. The llnravaged charaeter of
Ihe German assemblages seemed not in dispule.
ThllS Ihe difference in composiUon between Ihe
assemblages could derive from destructive agents
operative on Qoe and nol operafive on the olher as-
semblage. This was nol considered, Ano{her dif-
feronce not considered hy Read-Martin and Read was
the argument that Ihe original compoetuon of the
Makapansgat essemblage had been complete skele-
tons. The deposilional character of the German sttes
was nol that of complete anirnals. bul accordlng to
Read-Marln and Read (1975) was a biased aban-
donmenl of parts near a kili sile coupted wilh parts
consumad whle the hunttng party was living there.
This difference ulone could account for frequency
contrasts. but t was not considered. Read-Marln
and Read took the differences batween Makapansgat
and the German sues at face value snd considerad
e8ch lo be refp.rableto a differenl sel of behaviors on
Ihe part of {he human-hominid occupants:
inslead of assuming Ihal ltnimllls sC8venged fram
hominid kiJ/s. ir is hypoth6sizerl Ihal hominids
sCllvenged fromIhe killsaf Cllrnvores, a bellarand more
comprehensive explanolion of Ihe MIlKapllnsgat re-
moios. and one more in );eeping wilh Darl's workand
with ocologlral principIes.. scavellging australo_
pithecines brought back the lightest and mosl val-
uable bovidremains in lerms of their meal, mllrrow, or
polenlial use as 10015. Cranial fragmenls are overrepre-
senled because these are lhe parls mosl afien efl by
carnivoresand byscavengers such as hyaenasand vul-
tuces IRead-Marlin and Read1975:3631
This was a case of atlempling lo reach conc!usions
about the pas! in the absence of a melhodology.
Following Read-Marln and Read's publicalion, 1
presented a considerable body of informalion based
on elhnoarchaeological study of fauna among Ihe
Navajo and among the Eskimo (L. R. Binford and l. B.
Bertram 1977), My \\'ork in both places c:onfrmed
Ihe impressive abilily of gnawing dogs lo deslray
bones and therefore lo distorl Ihe relalive frequen-
cies of anatomical parts so Ihal the surviving popula-
Iion bore liule resp-mblance to Ihe composition Ihat
existed prior to Ihe section of the dogs. 1carried re-
search 00 lile differenlial slrength of bone a step be-
yond Iha work of Brain Bnd showed a strang rela-
tionship between bone density and the survival of
bonf'.5 gnawed by dogs. II was further showed that
bone density varied with lhe malurational age of
animals: tnerefore, the characlerislic profile of bones
SlJrlriving after having been by Ihe same
dogs wil1 be vastiy differenl if the prey acted llpon
are of differenl ages. One cauld Iherefore expeel to
observe r1ifferences in Ihe relalive frequendes of
analomicol parts survvlng even if all behevcrat
conditlnns were Idenrir-al except for the age compo-
sition of Ihe animal popuiation being ected upon by
altrilional agents.
Using an eleborate curve-fitting slrategy that
sought the best fu among a series of age-variable
popuiattons, we were able to account lar apprnxi-
rnateiy 79% of the varance in the Makapansgal
population. At Ihe time 1was imprcssed, anri tended
lo view the fit as reasonable, particularly since the
Makapnsgal population was made up of a variely of
species and our models had been based exclusively
on sheep. Accepling the unaccounled-for variance as
"noise," r viewed Braio's argumenl thal Ihe original
composition of Makapansgal fauna had been com-
plete animal skeletons thal had becn ravaged by
predator-scavengers as reasonable. This was essen-
tial1y where Ihe African situarion had been left re-
garding the inlerpretation of faunal assemblage var-
iabilily from early hominid siles until the presenl
Although the "eady" materials were largely left
alone during Ihe late 1970s, there was 8 stepped-up
series of comparatve studies done on more recenl
materials. Richard Klein has syslematically sludied a
large number of faunas from importa nI archaeologi.
cal sites in Soulh Africa (Klein 1975, 1976a, 1976b,
1IJ77, 1978a, 197Bc, 1978dj. He has made sorne in-
novalioos in faunal analysis and in tum has recog-
nized sorne interesting palterns. Of particular inler-
est are the differences he consistently noled belween
Ihe surviving analomical parls of animals of different
size. This pattern was firsl ciled in his analysis of
fauna from Ihe Klasies River mouth siles (Klein
1976a). Klein noled rhat among small and small fa
roedium bovids scapulae, pelvic parts, mandibles,
and parts of !he upper legs both fronl and rear were
commonly represented in the sites, but vertebrae and
lower leg parts were less well represented. This
makes perfecl sense in terms of economizing argu-
menls slressing Ihat Ihe mosl usable parts wiII be
Iransported to living siles and parts of marginal util-
ity will be differentially discarded al kill siles. If
such ralional behavior was manifest with respeel to
the small animals we could expecl it to be even more
obvious with large animals, which would present an
ellen grealer Iransport problem. Klein has syslemati-
cally observed jusllhe reverse: "the ralio of crunial
' dlI
vast body of behavtorally controlled material (L. R.
Binforcl 1978b} that should leave little doubt that the
model of Ihe past that admilted only residential and
kili sitas was extremely nalve. l am of the opinion
thet the beginnings of a methodology have been
worked out for identifying stte functons. at leas!
with regard lo hunting logisucs. We need at least as
much research investment in the behavioral cense-
quences of anmal-generated feunal assernbleges.
I have focused on assemblage variability consid-
ered excJusively in terros of anatomical port frequen-
cies. Clearly this is not Ihe only property of as-
semblages. We could study assemblages froro the
perspective of their frequency composition in lerms
of different species, different ages of the different
spedes, or a number of other zoological properties.
Sludies of this type have been conducted for Blong
time. Tile composilion of assemblages in lerms of
species frequencies is mosl commonly interpreted as
a reflection of environmental eomposilion. Much of
the current work in taphonomy js concerned with
the degroo lo which paleontological assemblages.
even assemblages referable to Ihe actions of a single
species. are representative of the frequencies of
species in the surrounding habitat (see Hehrens
meyer and Dechant-Boaz 1980, Brain 1980, and
particularlyGifford 1981). There is, however. another
possible conditioner of species frequencies, and
Ihal is Ihe bias of predator-scavengers for exploil-
ing prey niches ralher thBn taking a random sample
of prey in the habitat. Important work of direct
relevance to Ihe problem of identifying differenl
predator-scavengers products has been done by
Vrba (1975, 1960). af very great importance is a
provocative observation by Richard Klein (1976a,
1980)lhal carnivore species. in terms of both num-
ber of species and numbar of individuals, Ilfe much
more common in animal den assemblages IhBo
in other types of eilher human or paleontologicBI
assemblage. This supports my observations that
animal-produced pseudotools claimed from Euro-
pean sites are more common from levels and sites
Other Types of Assemblage Variability
Other Types ofAssemb/nge VoriobilHy
role of differential preservetton through his study of
contemporery Hottentot peoples and thelr faunal de-
bris. Simultaneously he studied the behevtcr and the
fauna! consequences o African anrnals. initially
leopards (Brain 1968) and later porcuptnes (Brain
1980). Brain not only worked lo obtain recognncn
criteria for antmel-generated assemblsges. hut elso
moved in the dtrection of investigaling the causes of
differenlial bone preservation. Because there was a
large segment of the Africanisl communrty that gen-
eraJly disagreed wilh Dar\. there was an emphasis on
the study of enimal-produced essembteges. and this
bias continues although not so marked as perhaps
during Ihe early 19705.
Vou might say that my work and the work of sorne
Near Easlern researchers {perkins and Daly 1968) has
been largely "spin-off researeh" in Ihal the original
problems thBt prompled Ihe research on Ihe Ameri
can Plains, ami in Africa, were not Ihe majar
motivators for either my work or thBI of olhers who
began to explore faunal assemblage variability in
other areas. Certainly my own work was motivalBd
by the implications of the Americanisl research. For
instance, when I went to France in 1968 to study the
faune ol Combe Grenal, I had a vBgue idea Ihat site
functions should be manifest in faunal assemblage
variabilily. This WBS a projection of the Plains argu-
ments regarding kili versus residential siles. J ar-
rived in France with a copy of the then very new
study by Kehoe (1967}. I hoped Ihat I could use
{aunal variability to inform me Bboul the causes of
lithic assemblage variahility. It took me sorne time lo
realize thBt we had insufficienl knowledge as to the
causes of variBbilily in faunal assemblages. and
could nol use Ihe palterning seen at Combe Grenal as
a methodological device for learning about the be-
havioral contexts in which the stone 1001 as
semblages were generaled. It was at that juncture
Ihall ("..ame to adopt many of the positions presented
in Chapler 2, namely Ihat middle-range reseBrch was
crucial and Ihal we hBd to do it ourselves, since no
o\her {eld sougbt to explain facts of Ihe archaeologi-
cal record. In addition. ir was cJear to me thal ex-
plBining Ihe facts of the Brchaeological record con-
sisted of understanding Ihe dvnamic conditions Ihal
produced those facts. In short. we had to do !lthnoar-
chaeological and experimental studies. I went to
study Ihe Nunamiul Eskimo in 1969. We now have a
5. ccmccenroo: Patterns ofAssodotjQfl Stemmmg from the Behuvinr 01 Mon versus Thut 01 Becst
termine what scevengtng actuetly looks Hke when
seen in fauna! frequendes, we might gain a very dif-
ferent piclure of man's adaptatlon to the African set-
A.s can be seen, we have had essentially 30 years
of research treating assemblege variability in
anatomtcel part frequendes. Much of this work has
proceded reletively independently in Mrica. in the
Norlh American Platns. and to sorne extent in the
Near East among workers tnterested in the origins of
agriculture. Work started with essenhally two quite
different paradigmatic assumptlons- whtte's ptoneer
work assumed thal assemblage variability was refer-
able to lo) differential butchering practices, (b) dif-
fecenbal transport or abBndonmenl of anatomical
parts between kili sites and residentiallocatio
{el Ihe differential traflsport of sorne selecl parts ior
use BS tools. In Africa, Raymond Darl viewed the
differential frequencies of anatomical parts as having
derived from the biased introduclion of bones into
living sites for use as lools. He too recognized a basic
dichotomy between residential ites and kili sites,
bul his majn focus was on residential sites.
In Americanist studies Ihere was a growing
awarenesS that preservation was no\ uniform with
respeet to bones, as had been assumed earlier, but
was in faet differential. That is, aUbones subjected to
the same destructive conditions wou!d not survive
equal1y. lt should be noted that this "awareness"
arose through observation on contemporBry dogs
(Guilday), through consideration of possible conse-
queTlces of ethnographically documented behavior
(Pillaert 1969), and as a result of direct ethnographic
observBtion (Lyon 1910). In addition there was arel
atively quick accumulation of a compBrative library
of archaeologically investigated sites illustrating
thal the paUerns were highly variable. There was
sorne exploralion of the idea that the simple
dichotomy, between kili site and residential site WBS
perhaps somewhat naive (Wood 1968); and then, re-
search essentially stopped on this subjecl.
In Alrica there was a very differenl pattern. There
was no initial rush to comparative study. For many
the Makapansgal fauna remained unique. Gis-
cussion centered around the degree to which other
animals could have produced Ihis fauna. In Alriea
the work of ona man provided the model for re-
search. C. K. Brain sought lo gain control over the
lo postcranial parts ncreases whtle the ratio of ltmb-
bones to foot bones decreases with tncreesed sze of
bovtd [Kletn 1976a:87\." Klein "tnterpreted" this
phenomena by appeal to the schlepp eUect of Pero
kins and Daly l\968], while pointing out thet in hs
sttes the people aleo increased the number of skulls
introduced at the same lime they wete introducing
feet and not introducing pars with substanlial meaty
yield! Another nterestlng fact ebout the large bovids
at the Kla:sies River mouth sites is that they are
mostly Cape buffalo and the extinct giant buffalo,
both of whtca are predorninantly represented by
very young individuals and individuals of advanced
age (Klain 1976a:83). in terms of contrast, the bovids
of small lo medium size ware almost exdUsively
prime adult individuals. Bolh patterns-skulls aud
feel from young and old large animals and the
meat-yielding bones froro hovids of small to medium
body size-have now been documented for Nelson
Bay Cave (Klein 1977:23), Border Cave (Klein 1977).
Boomplaas (Klein 1978c). Buffelsk!oof (Klein
1978d). es wel1 as Klasies River (Klein HI76a).
Klein's persistence in analysis and study has re-
sulted in sorne very important pattern-recognition
work (well surnmarized in Klain 1980). What these
pattems mean has been largely interpreled by Klein
from arguments of others, sucb BS the schlepp effect.
or his Bttemp1 to recognize differences between kill
and residential sites. Alternativa srguments and at-
tempts to warrant viewpoints through arguments
from elimination have nol characterized mueh of
Klein's work 1hus lar. For instance. sorne of the ideas
advaneed by ReBd-Martin Bnd Read largely derived
from comparative ethnogrBphic study should
perhaps be investigBted. They point out that a model
of scavenging is no\ inconsistent with documented
behavior of modern hunters (Woodbum 1968: J. E.
Yellen, personal eommuniClltion). Not only is it
plausible, but the scavenging of kills by olher pred-
Btors ""ouId roughly limit Ihe uSBble parts lo those
that tended to be abandoned by otber predators, ren-
dering such behavior recognizable. As we will sea,
these parts are nol uncommonly heads and lower
limbs {Read-Martin and Read 1975:3(3). A model of
scavenging is also consislen! wilh lhe age bias in the
large bovids tmany very young and very old indi-
viduals). AH in BlI, a scavenging model makes more
sense Ihan does the schlepp effect. 1f we could de-
#' 4
Observations of Wolves and Their Behavior
the btased abendonment at the site of bones from
parts mosl for consumpon. The as-
seroblege tberetor looked rnuch like an animal kili
assernblage where consumpton had been on the
spot. Hill's compersone made iI appear that mnu
behaved much like antmals.
The "search" Ior typical or ontologcally cnarec-
tenstic asscmblagse when there te lttle understand-
og of what cnndirfona ther form, other than a vague
suspclon thet iolrinsic properties of the egent are
responsible. ts just a sltghtly modied version of lhe
Iradilional archaeological assumplion that var-
iabilily is referabl e to differences in Ihe persons or
agents producing the assembJages. Thal is, as-
semblage differences are referred to elhnoeultural
differenees among meno Things are not Ihal simple.
We mus\ understanrl the proeesses thal condition
differences and similarities before we may give
meaning lo them reliably. This is another way of
saying whal was stressed in Chapter 2-
middle-range research is of necessity rooted in ac-
tualistic sludies, studies of dynamic phenornena.
This normally means studies of conlemporary living
1 have already studied in delail a tremendous
range of variahility as it was generoled by the
Nunamiul Eskimo (L. R. Binford 1976bl. Although
Ihe variability produeed by mHnin this case appears
almost overwhelming, there was patlernil1g and
there was sorne redunrlancy in Ihe types of paUern-
ing produced correlated wilh the Iypes of setting. !he
structure of the site. and a wide variely of olher con-
ditions. In that sludy. I was inlerested in Ihe
dynamics of the behavior Ihal stoad behind Ihe as-
semblages. In a similar way it is imporlant lo begin
the laslt of understanding, from a behavioral
perspeclive. lhe slructure of bone popu!ations pro-
rluced by animals. As a step in this direction. I will
presenl summary informarion on Ihe bchllvior anrl
trealmenl of prey by wolves.
The ObSl!rvfltiollS presented here wP.re made duro
ing a field invesliga!ion among the Nunamiut Es-
kimo. illland huntl'Ts of carihou. The Nunamiul Uve
fll the drainHge diVIde of !he Rrooks Range of norlh
1956 for rnview] Irealing the degree to which ani-
mals. epectcatlv hvenas, eccurnulete bones al a den
was clone by Hughes (1954). He observed a den exca-
valed in a soft substrate in whch no bones or fecp.s
were depositad. This reseerch was really done in the
cnntext of answering the question as lo whether
animals actually accumulated bones in thelr lairs
end denso As late as 1970 the accuracy uf lhe claim
that animals actually eccumuleted bones was still in
doubt. Sutcliffe's work (197U) seemed to pul this
questton lo rest wilh an unequivocal yesoSubsequent
research has been more oriented 10ward obtaining
samples of animalgenerated faunal populatlons.
Andrew Hill collected faunal remains from two Afri-
can locales (HilI1975). Shipman and Phillips (1976,
1977) made collections from slill anolher area. Diane
Gifford has iniliated time-series studies of carcasses
lo determine wha! Ihe sequence and inlerrelation-
ships between disarlkulation ami bone modification
might be. These sluoies have nol }'el been com-
pleted. A number of hyena den sturlies have been
conducted (Henschel el al. 1979: Owens and Qwens
1979) ami we look forward lo more that art! currently
under way by HiIl and by Bunn (Glynn Isaac, per-
sonal communicationj. Faunal assemblages have
been eollected by Behrensmeyer and Deehanl-Boa7.
\1960) bul have not yet been reporled in terms of
analomical part frequencies. Alfhough we can cer-
tainly Jook forward to a large body 01' malerials in 'he
Mar fulure. Ihe only sludy J,lublished lo dale thal
addresses Ihe problem of recognizing the behavior of
animals as opposed lo thal of humans is Ihe work of
Andrew HilI.
In his thesis, HiIl (1975) sought to identify as-
semblages as Iypical of human behavior and conlrast
Ihem with assemblages assignable lo animal be-
havior. In Ihis he chose whal was at the lime one of
the few widely available descriptions of a fauna! as-
semblage dearly referahle lo human behavior, thal
rom Olsen-Chuhbuck reponed by Wheat (1972).
Wilhout understanrling what made it dislinctive,
archaeolog(sls accrilp!ed jI RS ontologically "human"
in ils composilion ano othar characlerisljcs.
Ironically. the Olsen-Chubbuek Site is one where
(a) there was a mass overkill and In) the animals
were proeessl'Jd for !ranspor\ by removing Ihe meat
from the bones. These condiliol1s ensureo Ihat Ihe
mea! was removed frorn !he parls nI' ulilily
and lowest weight-to-surfa;e-arr!Hralios. resulting in
Observcuoos o/ Wo!ves and Their Hehuvor
nlerest in assemblages generated by animals cer-
tainly grew out of Ihe eonlroversy surrounding
Raymonrl Darl's work. An early argllment aboullhe
Makapansgat fauna was Ihal il had been produced by
hyenas. One 01" the carliesl "rcl:enl" sturlics {sr.e Dart
Studies of Assentblage Composition CBused by
The hypolhesis is Ihlt pre_agricultura\ man will
have resources in a "natural," Ll!.,
fashionand thallhBincre8sed control over his resoun:eS
given by domeslicalion a\lows plannetl cropping,
which will be seen in Ihe formof exploitation patlems
which devialefrom Ihe nalural rannomone llliggs and
arman 1912:7).
Despite the fael Ihal the proposition is phrased as
a hYPOlhesis, no attempl has been really made lo lest
i1. It is like the hypothesis of Bonnichsen (HI
garding spiral fraclure; it is in fad a convention used
to ascribe meaning to pll.tterned observations made
on Ihe archaeolngieal record. The use of l\lis conven-
lion led the lale Erir. Higgs to recognize evidence for
domesticalion in the Paleolithic of Europe! 8uc:h
conclusions were reaehecl because there was no
understanding regarding Ihe faclors tha! con-
ditioned Ihe differenlial frequencics of animals of
variable age appearing as debris from hunter-
gathercr adaptalions. In lhe abscnce of undersland-
ing of causal processes. many myths are generaled. 1
have made no attempl lo survey Ihis Iiterature. al-
thollgh il is extensive and in need of critical review.
1 will reslricl furlher dis(;ussion to assemblage
variabilily conceived in terms of analomical parl fre-
quenci fJs.
5. Assemb!0l!c Compositinn: PIl!l('rns f Associllljull Stemming frum the R('hnvil>r nf Mon versus Thlll of Beus!
hominirl bchavior as npposed lo that of other anl-
mals connues lo eusure that our vtews of men's
past are largely dtstorted mylhs.
Paunal asserobleges have been studied in terms of
the relative frequeneies of animals of different ages.
This has bee a common strategy arnong researchers
sooking, lo recogntze domeslicated as opposed lo
wild animals exploited by anctent meno These
studes [sec Luapfin 1969, 1971: Duces 1969: Hggs
and larman 1972) all make a basic assurnpton re-
garding the exploitatlon strateges of hunters:
where carnivore remains are extraordinarily corn-
mono This appears to he true of the famous site of
Choukoutien, and probebly apples to other such 10-
catons as wetl. For nstence. the femocs burial cave
of Teshik-Tash {Movius 195:11 appears lo have been
an animal den for a considerable period of lime. To
what degree are the evidences of ritual cited at this
slte referable lo the behavior of anmals" The wide-
sprcad use of European caves by nonhominid
predator-scavengers ls well illuslrated by lhe iol-
lowing genenlization regll.rding Europe<Jn siles:
"The nalure of Ihe site controls to sorne exlent the
nature of the fauna. Cave siles, far example, are often
rieh in carnivore remeins out of aH proportion to Ihe
abundance of the ereatures in the wild IEvans
In fact. two discussions of hunter-galherer sub-
sislence settlement strategies during the Paleolithic
ofGermany were almosl certainly .oled astray" by Ihe
old assumplion of Ihe monument and relic era Ihal
things found in lIssodation with tools showing the
hand of man were siso referable lo Ihe acHons of
meno Gamble (1979:39) aceepts the numerous pred-
ators (sometimes representing 45% of Ihe faunal re-
mains exclusive of cave bear) as largely the prey of
human hunlers. He noled that siles wilh high fre-
quencies of predalors tend to be small caves and
roekshelters that a'60 generally yield rele.tively small
artifacl assemblage;. He eontrasts such sites wilh
large arehaeological sites in the region lhat are nu-
merically dominated by herbivore remains. in an al-
tempt lo obtain a "regional" perspeetive on the hunt-
ng stralegies of andent manoIt is very Iikely Ihat he
is primarily contrasting the behavior of earnivores
wilh the behavior of meno
In anolher study Gamble (1978) is concerned with
the ehanges in both resource utilization and setlle-
menl Ihrough a time sequence. He clearly demon-
strated thal there is a major change betwccn Ihe
Middle Paleolithic and early Upper Paleolithic seen
as a unit and the subsequenl Magdalenian in pal-
teros of faunal association. The sites of the earlier
time period have large numbers of carnivores,
whereas Ihe Magdalenian has very few. Gamble con
sidered Ihis to be a measure of human behavior. I
strongly suspect that il refers to the patterns of in-
teraction between eompeting predalors-man and
Ihe region. Our inability, or in manv cases
our faHura to try, lo recogrdze thu cOllsequences of

,,_ i.N'mll/uA" CrJlJlf",<;JIIOW 1'(111"'''' nI i\.'s,wiulioll SI"mmiug mm rl", 1I,'hrJViof ,,1 M(l1l \ "f,'IIS 'I'11<I! n( flf",-,I 1>1' W"lves (!!ld Thf'lr Hl'havior
H(aIRE !UIl. Wu/ves r,,/oxiIlR(I! pninl.
central Alaska. Majnr segments of the An:tc r-aribou
herd bannuattv move turough ths area from their
wnter range in the foreste lo l!lR south, lo the sum-
rner feeding arcas on the tundra \0 rhe north (seo L.
R Hlnfonl 1!J7Rb:l:l. Figure 1.1). Mus! of the wolvcs
o the (;l'utral Hrooks Renge nre datk hr own 10 LIad;
a lesser numher are gray. and 1 have obsurvud an
elbruo WlJJr un sevcral occastons. Of Ihose killed hy
the Nunamiut dunng the winter of ]970-1!"l71 [five
uidividuals. three males and Iwo femalesl. the aver-
age weight was /J7 lbs. Dunug the fieldwork in and
around the Anaktuvuk Pass atea. I spotted 23 single
wolves and 8 packs. Of the packs obscrved the muan
size was fi.:m animals with the smallest group being
2 animals ami the largesl 11; a total of 51 anirnals
with the smallesl group being 2 anlmals ami the
largest 11; a total ol 51 animals were observad in
packs. I observed onlv ono Kili by a pack. but cerne
upon feedng wolves 011 twu otber occasions. I had
the opportunity to examine in sorne detall two wolf
dens and made ubservations on :Uwolf kilI and feed-
ing siles.
Dens and Loirs
In earlier sludies almosl any repealedly used loca-
\ion was caBed a den_ Fur instance, what Hughes
(1954) called a den was aclual1y a nest area wbere
young pups were bom ami kept while lheir eyes
were slill ciasen. I will rontinue lo use rlen lo refer lo
Ihis type of sil!'._ However. 1 musl dislinguish be-
Iween a den and a oir: the lalter is the (;Cnlrlll place
where yaung 10caJize lheir aclivilies and where
adult wolves spenrl most of Iheir lime adjal:enl to a
den. Lairs generally consist of two arcas. a traHie
rea and a defe(:alion urea; thl:' olJer animals tend to
muve away from the cure activity mne of the lair fm
defecatiun. In t1wBrouks Rallge of Alaska most
are dug into SUlllh-ffldn;: cut banks and slopes; hnw-
ever, Ihere are a fe..... rm:kshelters and caves in 1P.
Iimestone lhat occasionally serve as denso In lhe im-
mediate area of Anaklllvuk, mus! (:aves or rocbhel-
ters are quite high along lhe mountain slopes and th!!
Eskimo report thal wolves prefm lll"wer place whl!m
they can "run" out inlo lhe vallev Iw\tf'T"
Hobl'rl (J, who has (;lJndlll:ll!d
si ve Slllrlips oi wolves in tlw AnakllJVuk fl'-
aod m_v diw('\ obsvrVillillns as wl.'II11S lIf
f ntonnauts confirm. lhal wolves dig burrows (ur un-
largc other animal burrows in sand and day and
make use of fisstlres i 1lstono subst ratos fnr 1heir dens
IStephenson <I11d tohnson 1H72:l!J\ [len collapse is
rornmon in sand. Fromoverly liry condifions in the
ama. and in clev from oxcesstve mnistun: (Stephen-
.'lOO end )oh115oll resulfing in only Iimited
reuse of a gtven den. This is nbviously not thc case
for rnckshulter, cave. and c.rcvk:e denso which mav
be reoccupicd almost endlc-slv: "Naturally uccur-
riug rock formatkms are the ml;sl stahle substrate in
which wol!' dens occur anrl. in sorne cases. have pro-
vded e....senually unchanging den sitas for decades
aud perhaps even centuries ISlephellson and
lnhnson 1972:2YI'
The actual pauorn of reuse of a given den was in
the opinion of tbc Eskf mn nol conunuous: Ihat ts.
dens were beueved lo be reuserl rm un evcrv-other-
vear bass by the wolves. Inrormants loirl both
Stuphcnson (Stephenson and Johnson: 1972;21) and
me thet ths was because of lhn lngerin;: smell from
lhe earlier year; "Ihey wail" tilllhe smell ;oesaway.
The females are consklered by the Eskilllo lo have
"deoned up" by the firsl week in May. Each rlen is
consirlered lo be "Hile HfHmily" with a single malure
a mature Illflle. ano a variflble numbnr of im-
malure animals, Activilies of lhf1 family are helieved
lu be centp.wd on lhe lair throu)!.hml1 lhe summer.
Only \'Vith the fall migmlion nf Ihe caribou are lhe
families believed to come logelllf'r inln lflfger packs
thal move loward the limberline behind lhe caribull.
Wolf lairs are. then. highly seasoofll and specilic lo
spring and early SlIlllml'r As the pups be;omr-
larger, the adults may increase lheir ranginR in
search of and may lakl' Ihe pups "out.' as the
E..imo say. leflving lhem in a lhicket. dilch. or other
local ion while the adulls travel over lbe
terrilory h\lllling. mturning 10 Ihe pups either to
move lhem agflin