See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/230230080

Of taphonomy and zooarcheology. Vertebrate taphonomy. By R. Lee Lyman

(1994)

Article in Evolutionary Anthropology Issues News and Reviews · June 2005

DOI: 10.1002/evan.1360040209

CITATIONS READS

53 376

1 author:

Curtis Marean
Arizona State University
215 PUBLICATIONS 15,271 CITATIONS

SEE PROFILE

All content following this page was uploaded by Curtis Marean on 20 February 2023.

The user has requested enhancement of the downloaded file.

64 EvolutionaryAnthropology

Of Taphonomy and
Zooarc heology
Vertebrate Taphonomy
By R . Lee Lyman (1994). Cambridgc:
Cambridge University Press. 576 pp. $79.95
( c l o t h ) , $34.95 (paper). ISBN
0-521-45215-5 (cloth), 0-521-45840-4
(paper).

In the last twenty years there has

been an explosion of taphonomic re-
search with a zooarcheological bent.
Africanist and North Americanist
archeologists have conducted much of
this research, stimulated by debates
about the interpretation of African
and North American data. With regard
to North America, these debates dealt
with the related problems of the settle-
ment of North America, Pleistocene
overkill, and hunting of extinct mam-
Figure 1. Lion eating a Wart hog
mals. In Africa, the focuses were Dart’s
interpretations of faunal assemblages
that is not a deficiency because this mammals. Good critical discussions
from the South African breccia sites
group has done most of the research. of important research fill the book
and Glynn Isaac’s home base m o d d
The book h a s a n unabashed and, in many cases, are enlarged by
Much of this research has had a parc-
zooarcheological flavor that is the re- Lyman’s own ideas.
chial quality; the North Americanists
sult of Lyman’s grounding in A work of synthesis such as this al-
have rarely seemed aware of the A f r -
zooarcheological problems. This is a lows one to stand back from the trees
canist literature and vice versa. Unl,il
strong plus for the book: ultimately, and see where additional research is
now, there has been no single treat-
the questions we ask of the fossil re- needed to fill out the forest. In this re-
ment of this geographically disparate
cord are what make all taphonomy view I will identify three areas tha.t re-
research.
relevant, and most of the questions quire such growth, which this book
In recent years, taphonomic re-
guiding recent vertebrate taphonomy unintentionally highlights. I say unin-
search has reached a finite mass and. a
have been zooarcheological. In the last tentionally because Lyman does not
new level of maturity. Consequently, it
several years there has been a melding identify these areas as being deficient
has warranted and, indeed, begged for
of taphonomy and zooarcheology, but or problematic. This book is a gift to
synthesis-a treatment that would not all zooarcheologists but, as with most
only gather together the major results, there has been no single book that ade-
quately covered the material. I teach gifts, there is something more thai one
but also elucidate the guiding themes wishes was part of the package. My ex-
that bind these studies into a scientifi- two courses on zooarcheology and
both have lacked an adequate text- tra wish list includes a more critical
cally meaningful whole. This ta.zk t r e a t m e n t of the structure of ta-
would require an individual with ex- book. This volume fills that gap.
phonomic epistemology, a call for ta-
perience in primary taphonomic re- The book displays Lyman’s typical
phonomic research joined to studies
search, a s well as in describing a scholarly fastidiousness. I found no
of prey mortality, and a stronger state-
variety of zooarcheological assern- omissions or errors in citations of the
ment regarding the critical signifi-
blages. Of the few people who could literature I know best, and that is an
cance of the study of shaft fragments
have been up to the task, R. Lee Lyman enormous accomplishment for such a
to zooarcheological methodology
headed the list. Thus, all scientists broadly based book. Moreover, the
who are interested in taphonomy or book has several characteristics that
make it essential for anyone interested TAPHONOMIC THEORY AND
zooarcheology can rejoice at the pub-
lication of Vertebrate Taphonomy by R. in taphonomy. Using a standard for- DIFFERENT TYPES OF
Lee Lyman. mat, Lyman repeats the original data TAPHONOMY
Lyman provides an even-handed from many separate but related publi- As taphonomy has grown in popu-
but critical review of the vast majority cations, including, for example, the larity, t h e number of researchers
of vertebrate taphonomic research important data on bone density and claiming to do it and the number of
conducted up to 1992. Most of the lit- bone utility. The book covers ta- their studies have both increased. The
erature covered is in English and was phonomy of all types, including that result is a plethora of views on what
done by North Americans. However, pertaining to fish, birds, reptiles, and taphonomy is, how it should be done,
BOOK REVIEW Evolutionary Anthropology 65

and what central principles should Actualistic studies result in generaliza- ous epistemology, I believe it is critical
guide the structure of the research. In tions about the relations between to separate the two and to recognize
the first three chapters of this book, processes and patterns. These gener- the distinctive strengths and limita-
Lyman provides a comprehensive alizations become our tools for inter- tions of the different approaches. To
overview of the topic, discussing the preting the meaning of patterns in the understand this necessity in full, we
nature, history, a n d theory of ta- fossil record. need to summarize points made in
phonomy He argues that the basic In chapter 3 , Lyman summarizes Gifford-Gonzalez’s3recent discussion
goals of this discipline are to remove three examples of taphonomic re- of taphonomic theory, which 1 believe
the taphonomic overprint from the search that he considers exemplary, to be the most significant yet offered.
fossil record in order to obtain accu- presumably because they follow the Gifford-Gonzalez3builds a system
rate resolution of the ancient biotic goals and procedures just outlined. of analytical categories that goes a
community; to identify the character These studies begin by developing long way toward solving many termi-
of the taphonomic overprint in order methods, or taphonomic models, for nological ambiguities in archeological
to reconstruct the precise processes interpreting the fossil record, then ap- inference. Lyman discusses this sys-
that resulted in the fossil record; and ply these methods to patterning in the tem in detail. A trace is a visible attrib-
thus, finally, to reconstruct the ta- fossil record. All three examples are ute displayed by a bone or any other
phonomic history. Attaining the sec- archeological find that has undergone
ond of these goals often results in data a taphonomic process (e.g., a scratch
of paleoecological significance. Gif- on a bone). The causal agency is the
ford,’ in her review of the relation be- immediate physical cause of a trace,
tween taphonomy and paleoecology, Researchers have the such as a tooth scraping across a bone.
argues that the goal of stripping off the The effector is the item or material that
taphonomic overprint, or unbiasing choice of using a true effects the modification of a bone,
an assemblage, is an unproductive actualistic model, as such as the tooth. The actor is the
and, perhaps, impossible methodo- source of the force or energy that cre-
logical agenda. She advocates an ap-
warranted by ates the trace, such as a hyena biting
proach that is more consistent with uniformitarianism,or on a bone. The goal of this analytical
Lyman’s second goal. what is commonly construct is to assist us in evaluating
The differences here may result the strength of bridging arguments
from the subdisciplines Lyman and called the comparative that link a trace to a causal agency, ef-
Gifford attempt to entwine with ta- method. fector, and actor. It is in this linkage
phonomy; paleoecology in her case, that we find the critical difference that
zooarcheology in his. Zooarcheolo- separates comparative from actualis-
gists, by definition, are interested in tic studies.
the behavior of ancient people. Thus, Actualistic taphonomy includes
fine studies, but they vary in explana-
they tend to see the diverse nonhuman both a naturalistic and an experimen-
tory power because they rely on differ-
processes that affect a fossil assem- tal component. In a naturalistic study,
ent sources for their taphonomic the analyst observes the occurring
blage as a bothersome fog that ob-
models. This variation, which is also process and the resulting pattern;
scures the composition of the
characteristic of zooarcheology as a there is no doubt about the relation
assemblage as originally discarded by
whole, is a source of confusion. Some between the trace and the actor. The
people. Zooarcheologists are not in
of the studies Lyman cites use the ac- analyst is not a participant in this
agreement on the proper perspective.
My reading of the zooarcheological lit-
tualistic method to create their “uni- process, which takes place in a natu-
formitarian umbrella,” others do not. ralistic context, and does not inten-
erature suggests that those who work
with food-producing and complex so- The strength or weakness of these tionally manipulate the parameters of
cieties tend to equate taphonomy with analyses rests on the nature of the ta- the process. A good example is
bias that needs to be cleansed, whereas phonomic model used to interpret the Blumenschine’~~~~ research on carcass
those who work with hunter-gatherer archeological pattern. From what consumption by large African carni-
societies, particularly those of pre- source does that taphonomic model vores. Blumenschine observed the se-
modern humans, have a more paleoe- derive its understanding of the pat- quence of consumption by the actors,
cological approach. tern-process relationship so critical to saw the effector create the trace, and
Lyman notes that there is no rule on uniformitarianism? then documented the causal relations.
how to meet the goals of taphonomy Researchers have the choice of us- Blumenschine did not artificially ma-
except, perhaps, the use of a “unifor- ing a true actualistic model, as war- nipulate the number of carnivores in
mitarian umbrella.” He argues that the ranted by uniformitarianism, or what any way, and did not otherwise inten-
uniformitarian approach is bound to is commonly called the comparative tionally modify any experimental pa-
actualism, and that actualistic method method.ls2 Some studies Lyman cites rameters.
entails observing present processes in rely on the comparative method, while The second component of actualis-
action, then establishing causal rela- others rely on a true actualistic model. tic taphonomy is experimental stud-
tions between processes and patterns. If zooarcheology is to have truly rigor- ies, which are similar to naturalistic
66 Evolutionary Anthropology BOOK REVEW
-
studies in that the analyst observes the ied only results in a false sense of If comparative studies cannot re:jult
relationships among the trace, causal greater confidence regarding studies in method because they do not meet
agency, effector, and actor. The differ- based on modern hyena dens. With the litmus test of actualism and uni-
ence is that the analyst actively con- comparative studies, several chains formitarianism, then what is the
trols the parameters of the observed are missing from the analytical linkage proper role of comparative studie,s in
process to increase understanding of established by Gifford-Gonzale~,~ par- taphonomy? Such studies are a fertile
the relations linking actor to trace. An ticularly the link between the trace and source of hypotheses that actualistic
example is my experimental work with its causal agency. The results of com- methods can test. In what I consider
captive spotted hyenas, in which I and parative studies are bridging argu- an ideal research structure, compara-
my colleagues intentionally varied the ments that rely on circumstantial tive studies result in hypotheses about
number of hyenas to simulate differ- evidence. I do not believe this is a safe linkages between trace and actor, after
ent levels of carcass c~mpetition.~,’We source for secure archeological infer- which a researcher designs a natural-
did this in order to understand the de- ence. istic study to test the validity of the hy-
structive impact of hyenas on archeo- Returning to my example of com- potheses. Following the naturalistic
logical bone assemblages. parative studies, we might ask what study, experimental studies further re-
There is a dynamic relationship be- potential inferential problems could fine understanding of the linkage:; be-
tween naturalistic and experimental arise? There are many. Suppose we are tween actor and trace. Ideally, the
studies: the former set the research using the relative representation of result is a robust bridging argument
agenda; the latter refine our knowl- skeletal elements from the hyena dens that can then be used effectively in ar-
edge of the bridging arguments. to establish a method of recognizing cheological interpretation. An impor-
Blumenschine’s research stimulated hyena dens in the fossil record, an ap- tant epistemological step has been
my own, and we have recently begun proach that is currently popular. The missed if a researcher moves directly
to integrate the results? relative representation of skeletal ele- from comparative studies to archew
The comparative method is very dif- ments becomes our trace and hyenas logical inference.
ferent. Explanatory models drawn the actors. Because no one observed
from comparative methods are the re- hyenas bringing in all the bones, the MORTALITY STUDIES IN
sult of studies of traces in which the possibility exists that other actors ARCHEOLOGY
relation between actor and trace has (jackals and leopards) contributed to Mortality studies typically investi-
not been observed, but often is be- the trace. Second, because the den gate the representation of different
lieved to be known. For example, sup- probably accumulated over a consid-
age groups of prey with the goal of in-
pose we take five hyena dens erable time, there is the possibility that ferring the faunal exploitation behav-
excavated in the last 30 years, study other processes removed bones from ior employed by the predator.
them, and arrive at inferred linkages the assemblage and thus systemati- Although studies of the season of
between trace and actor with the goal cally changed the assemblage. These death of animals could properly be in-
of developing methods to recognize processes could include bone weather- cluded here, I will focus on studie’sthat
fossil hyena dens as opposed to human ing, which destroys less dense bones, use mortality data to reconstruct the
bone accumulations. The linkage be- and porcupines, which remove large manner of death of the taxon of con-
tween trace and actor is inferred: no dense bones. cern.
one actually saw all the bones being The actualistic method produces In chapter 5 of his book, Lyman pro-
accumulated by hyenas. This infer- bridging arguments where real obser- vides a detailed and comprehensive
ence, then, becomes the bridging argu- vations connect all analytical levels, discussion of mortality studies, begin-
ment used to identify fossil but in comparative studies the connec- ning with the basic mortality charac-
assemblages as hyena dens. Taking tions can only be surmised. In actual- teristics of living populations. He then
this one step further, we may even add istic studies, the researcher discusses the typical models of mortal-
a fossil hyena den to the list of assem- confidently knows the causal agent. In ity used to interpret mortality pro-
blages used to develop the actor-trace comparative studies the researcher files. These include the catastrophic
connection. In fact, the comparative does not confidently know the causal model, in which the frequency of age
method as advocated by Klein and agency and, indeed, may have inferred groups resembles that found in a living
Cruz-Uribe2 involves comparison be- it from bridging arguments developed population, and the attritional model,
tween fossil assemblages only. How- in other comparative studies, Thus, in which the frequency of very young
ever, bridging arguments derived from comparative studies are missing a link and very old individuals surpasses that
comparisons between fossil hyena in the analytical chain; the link be- of other age groups. Recently a third
dens differ only in scale from bridging tween trace and actor is circumstan- model, the prime-dominated model,
arguments developed from compari- tial. Studies such as these do not meet which includes a dominance of prime-
sons between modern hyena dens. In the definitions of the actualistic age adults, has been discussed.9,’O
neither case are there strong controls method as outlined by Lyman because Lyman provides an extremely useful
on the relationship between actor and no one observed pattern-process rela- discussion of some analytic problems
trace. The difference between the tions. Archeology cannot afford to associated with the interpretation of
times at which the assemblages were have methods developed from circum- mortality profiles, then moves on to
formed and the times they were stud- stantial evidence. the theory of interpretation. Mortality
BOOK REVIEW Evolutionary Anthropology 67

studies typically examine a mortality ond assumption is that once the den- noarcheological research.15-18 Al-
profile, which is either a frequency titions or body parts used to construct though the results have not solved
histogram or ternary diagram display- the mortality profile reached the site, every problem, several useful patterns
ing the relative representation of dif- they were not differentially preserved. are emerging from them. One pattern
ferent age cohorts. Researchers can Klein14 has noted the problems of dif- is that, on average, hunter-gatherers
estimate these age cohorts by measur- ferential preservation. Lyman states transport small mammals more com-
ing the crown heights of molars or that the prime-dominated equid pro- pletely than they do large mammals.
from wear-stage classifications. The file at Solutre may have formed from For example, a single person can carry
resulting profile is the data used to in- the lack of deciduous teeth due to de- any animal up to the small end of the
fer the killing behavior of the predator, structive processes, but there are few size two category, such as sheep and
which in zooarcheology is often a discussions of this problem elsewhere goats (see Braid9). Mammals of this
hominid. The models typically used to in the mortality literature. The forego- size are sometimes transported com-
interpret the profile can have several ing assumptions are most likely incor- plete from the encounter site to the
sources, including general ecological rect; at the very least, they should not residential site. Any mammal larger
principles, comparisons to modern be essential parts of an interpretive than small size two is typically field
kill data, or comparisons to other ar- model. butchered and transported in seg-
cheological assemblages. Lyman cor- ments. A second pattern is that small
rectly notes that major progress has groups of people transport a carcassless
been made in the analytics used to completelythan do large groups. A thrd
construct the profile3 and in under- pattern is that transport decisions vary
standing the modern range of vari- Researchers typically by species and sometimes even by sex.
ation in mortality patterns.gJ0 The importance of these patterns
Unfortunately, there has been little treat the mortality profile with regard to mortality studies is that
progress on the taphonomy of mortal- as a perfect record of any one species could pass through
ity profiles. More specifically, there several size classes within its lifetime.
are few actualistic studies of mortality the relative For example, the wildebeest ( C o ~ n o -
(for exceptions see Hudson” a n d representation of age chaetes taurinus) remains a size one
Blumenschinelz). We know little or animal while it has its deciduous den-
nothing about the processes that re-
groups killed by a given tition and, when it begins to shed
sult in formation of the mortality pro- predator. This procedure these teeth, enters the size two cate-
file that is ultimately our source of is akin to treating the gory. By the time it reaches size three,
inference. This fact jumps out at the the wildebeest has a full set of adult
reader of Lyman’s book one moves relative representation teeth. If we were examining the mor-
through hundreds of pages of ta- of skeletal elements at tality profile of wildebeest (or zebra,
phonomic research devoted to the for- red deer, or bison, etc.) at a cave site
mation of bone assemblages, then an archeological site as and wished to interpret that profile
finds no comparable research on the a perfect representation with regard to the killing behavior of
formation of dental assemblages and the wildebeests’ predator, we would
mortality profiles. Researchers typi-
of the body parts used need to assume that there was no dif-
cally treat the mortality profile as a by a predator. ferential transport of the different size
perfect record of the relative repre- heads of wildebeest of different ages.
sentation of age groups killed by a In contrast, I would assume, based on
given predator. This procedure is akin the extensive actualistic data cited
to treating the relative representation To my knowledge, these assump- earlier, that, on average, the heads of
of skeletal elements at an archeologi- tions have neither been tested with ac- size one wildebeest (juveniles) made it
cal site as a perfect representation of tualistic data, n o r received any to the cave site more frequently than
the body parts used by a predator. Ana- detailed treatment in the literature. did those of size two or three wilde-
lysts must make two fundamental as- This seems clear from the lack of any beests (adults). The resulting mortal-
sumptions if they wish t o treat discussion of these problems in Ly- ity profile would be a reflection of
archeological mortality profiles as re- man’s comprehensive review. Because transport, not killing behavior.
flecting the killing behavior of apreda- these assumptions are directly con- Lyman cites Binfordszo useful ax-
tor. The first assumption is that the trary to much of the evidence pre- iom: “A transported assemblage
heads of juveniles and adults were all sented in this book, it is important to should look like the opposite of the
left at the encounter site or trans- have a closer look. nontransported assemblage; that is it
ported to a residential site without [the transported assemblage] should
bias toward any age group-in other be the proportional inverse of the
words, that none were subject to dif-
DifferentialTransport parts that were not transported.” Al-
ferential transport. Binford’s13 cri- The differential transport of large though Lyman cites Binford while dis-
tique of this assumption has herbivore carcasses has received a cussing bone representation, this
unfortunately been ignored. The sec- great deal of attention in eth- statement is equally applicable to
68 EvolutionaryAnthropology BOOK REVIEW

assumes that there is no secondary

9
butchery site between the primaq one
and the residential site. However, this
model is more complex and realistic
than the one currently in use. Once the
actor has discarded the transported
profile and it has entered the ground,
the profile becomes subject to forces
of differential destruction.

Differential Destruction of
Deciduous and Permanent
Teeth
Lyman provides an excellent discus-
sion of destructive processes and the
survival of bone in archeological de-
posits. These processes include bone
weathering, trampling by people ,and
other animals, destruction by carni-
100 % Old vores, sediment compaction, chemi-
cal leaching, and others. Actualistic
research has shown that some of these
processes are mediated by the fact i hat
less dense bones are destroyed more
easily, quickly, and frequently than are
more dense bones. In the 1980s, Ly-
man's21.22 research on the quantifica-
tion of bone density proved the
significance of differential density in
the interpretation of the differential
representation of skeletal elements.
Many studies have now documented
the differential density of bones of dif-
ferent species. Many zooarcheological
studies have used these data to seek
the impact of density-mediated proc-
esses on bone survival.
With bones Just teeth Deciduous dentitions and the alveo-
lar bone that encases and protects the
Figure 2. Mortality profiles from site GvJrn46, Nliddle Stone Age levels.Top, the profile as a frequency teeth in young animals are almost cer-
histogram.The dark grey bars representteeth only; the solid black line shows the inclusionof unfused tainly less dense than are permanent
bones. Bottom, a ternary diagram with the mortality models of Stiner9,10included. The filled star is dentitions and adult alveolar bone.
the position of GvJm46 with teeth only; the uifilled star is the position corrected for the inclusion of
Deciduous dentitions are lighter and
less well cemented, particularly in
neonates and very young animals.
mortality studies. Within a given sys- realm of actualistic research, an unex- This makes evolutionary sense: de-
tem of mobility, the mortality profiles plored realm for mortality studies. ciduous teeth are designed to last only
of encounter sites, where animals The mortality profile we are inter- a short time as compared to perma-
were killed or otherwise confronted, ested in is the death profile-this is the nent teeth. Further, a portion of the
should differ from those at residential actual number of individuals in differ- time during which a herbivore retains
its deciduous teeth involves suckling
sites, where animals were taken. All ent age cohorts killed by the predator.
rather than the consumption of vege-
things being equal, I would expect Butchery and transport decisions split
tation. 1 know of no data on the density
hunter-gatherers to transport the this profile into an encounter site pro- of deciduous versus permanent teeth,
heads of younger animals and leave file, that provided by what was dis- but a study in progress in my labora-
the heads of older animals at the en- carded at the encounter or primary tory should begin to rectify this prob-
counter site. There are, of course, butchery locality, and a transported lem.
other contingencies that could opor- profile, that of the items taken to and Deciduous dentitions, depending
ate on this pattern, perhaps even re- discarded at the residential site. This on the species, make up the first or
versing it, but this is precisely the is, of course, a simplistic model, for it second age class in a mortality profile
BOOK REVIEW Evolutionary Anthropology 69

illustrated as a frequency histogram not a perfect measure of mortality to quantify the frequency of skeletal
broken into ten percent age classes. In either, for unfused bones are less elements. Zooarcheological speci-
the ternary diagram, deciduous denti- mens typically are fragments of whole
dense than fused ones.I9 It is reason-
tions include all of the juvenile class, elements, which zooarcheologists do
able, therefore, to suggest that even
which, along with prime-age adults their best to identify with respect to
more juveniles were originally present
and old adults, make up the axes. If an t h a n is indicated by the unfused skeletal element and taxon. As a rule,
analyst examines a n archeological bones. This leads me to suspect thatresearchers then summarize the data
mortality profile and treats it as if it other prime-dominated profiles may in one of two ways. First, researchers
represents the discarded transported reflect destruction rather than mortal-
may tabulate raw counts per skeletal
mortality profile described earlier, ity. element. This unit is often called the
then the analyst is assuming the ab- As Lyman notes, and as Gifford' number of identifiable specimens
sence of differential destruction that originally stated, a primary goal of ta-
(NISP). Most researchers go further
would have biased the profile against phonomy is the reconstruction of theand attempt to estimate the number of
the deciduous dentitions. However, taphonomic history of an assemblage.elements accounted for by the speci-
most, if not all, fossil assemblages To date, taphonomy and zooarcheol- mens. These measures include the
have undergone some destruction. ogy have made little progress towardminimum number of elements (MNE)
Thus, I think it is safer to start, as this goal in mortality studies, primar-
and its derivatives, such as the mini-
KleinI4has done, with the assumption ily due to the lack of actualistic re-
mum number of individuals (MNI)
that destructive processes have biased search. The taphonomic hstory of a and minimum number of animal
the profile against age cohorts derived mortality profile will include at least
units (MAU). The resulting numbers
from deciduous dentitions. the encounter site profile, the trans-
and the patterns in those numbers are
The problem of differential destruc- ported profile, and the excavated pro-
then used to discuss broad issues such
tion is perhaps the most intractable of file. The excavated profile is separated
as butchery and transport decisions
the problems facing morality studies. from the others by its inclusion of den-
and other questions about faunal ex-
How does an analyst test for the pos- sity-mediated processes that have de-
ploitation. Lyman provides an excel-
sible effects of the destruction of de- stroyed deciduous dentitions. The lent discussion of all these issues in
ciduous dentitions? How does one points I have raised above cannot bechapter 4.
estimate the level of destruction? This resolved with the current data because Some specimens are easy to iden-
is precisely where actualistic research of the lack of relevant actualistic data.
tify, particularly the epiphyseal and
could be productive. One possible way I suspect, however, that our studies of
nonepiphyseal ends of long bones.
mortality profiles have given us more
Other specimens, such as the shaft
to begin is to examine the fusion of
information about differential trans-
fragments of long bones, are more dif-
bones, then compare the results to the
mortality profile derived from teeth. I port and destruction than about ani-ficult to identify if, as is most often the
mal mortality. case, they have no readily observable
attempted this procedure for an as-
These are primarily methodological
anatomical landmarks such as foram-
semblage of bones from the Late Qua-
problems, but an issue of theoretical
ina and tuberosities. One can readily
ternary site of GvJm 46 at Lukenya
importance arises from them. We perceive the potential range of vari-
Hill on the Athi-Kapiti Plains in
need to view mortality profiles within
ation between researchers in the level
Kenya. The open-air site, which has a
the context of a mobility system be-of their identifications. This variation
Middle Stone Age and a Later Stone
cause they are subject to differential
could derive from differences in expe-
Age component, appears to have been
transport based on decisions derivedrience or simply the amount of time
a mass-kill locality. Nearly all the
from economic concerns. For ade- assigned to analysis.
bones and dentitions there derive quate evaluation of the killing behav-
from a single species of extinct al- Until recently, it was standard prac-
ior of a predator, particularly a tice in zooarcheology to focus on
celaphine antelope. The Middle Stone hominid, we need to examine the pro-
Age mortality profile of the extinct an- epiphyses and the ends of bones to
files from carcass encounter sites and
generate estimates of the relative rep-
telope superficially resembles a residential sites. Without using a mo-
prime-dominated profile. Figure 2 resentation of bones in assemblages.
bility system approach, we cannot
shows this profile as both a frequency This had the obvious benefit that re-
hope to understand the killing behav-
histogram and a ternary diagram. Be- searchers could study extremely large
ior of the predator. With this ap-
cause the site shows significant post- assemblages in a reasonable amount
proach, we would bring mortality
depositional d e ~ t r u c t i o n , I~ ~was of time. Recently, this practice has
studies more in line with the mobility-
concerned about the possible destruc- come under fire, and the ammunition
system-based theory that dominates
tion of deciduous teeth, and thus ex- has come from several sources.
hunter-gatherer research.
amined the fusion of bones. The bone KleinZ4 published a report of a hyena
fusion data record the presence of den that clearly showed that the shafts
MIDDLE SHAFT FRAGMENTS AND
many more juvenile individuals, so if provided higher element counts than
ZOOARCHEOLOGICAL did the articular ends and epiphyses.
I included those data in the mortality
QUANTIFICATION Bunn25showed the same situation in
profile, the profile moved from a
prime-dominated one t o a cata- N o issue is more significant to archeological assemblages and ar-
strophic profile. However, bones are zooarcheolow than the methods used gued that hyenas had consumed the
70 Evolutionary Anthropology BOOK REVIEW

Proximal End mn
Experimental Kobeh Mousterian

[1T1
becular bone has more bone grease.
Nevertheless, because of this inverse
relation between grease content and
-

ml
density, I think it is safe to call this a

-I+ Proximal Shaft density-meditated process.

Lyman's discussion does not take
C I I I I I the issue to its logical end, so I will do
so here. The middle shafts of bones
from animals in all size groups sunrive
destructive processes better than do
other parts of long bones for two i-ea-
sons. First, middle shafts are more

,-& Distal End

mu
0 10 20 30 40 50 60
Minimum Number of Elements
0 10 20 30 40 50 60
Minimum Number of Elements
dense than are other parts of long
bones.*' This assures that, given even
the smallest amount of destruction,
middle shafts will provide higher esti-
mates of long bone abundance than
will long bone ends. Second, no grease
Experimental Kobeh Mousterian is trapped within middle shafts, mak-
ing them an unattractive food item.
-Proximal End Even when carnivores are interested
in gaining access to marrow, they typi-
cally break the shafts first, then leave
-Proximal Shaft the shaft fragments behind and s d -
low the ends, either completely orpar-
Middle Shaft tially. Because the ends of long bones
typically have bone grease trapped in
trabecular bone, carnivores, if they
are able, always consume long bone
ends and leave middle shaft frag-
-
Distal End 111 ments. Every actualistic study that
has produced data relevant to these is-
sues has demonstrated these
points,5,6,19.24
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Minimum Number of Elements Minimum Number of Elements This leads to the conclusion that
Figure 3. The number of tibia and metatarsals represented by different bone portions. The histo- middle shafts are essential to any ac-
grams to the left show the MNEs by bone portion after consumption (there were initially 50 bones curate measure of the abundance of
of each, some complete, others broken by hammerstone percussion). The histograms to the fight limb bones. But the importance ex-
show the MNEs by bone portionfrom Mousterian levels P, R. T, Y, and W at Kobeh, from size one and
two bovids and cervids. tends further, and Lyman misses this
key point. The actualistic data show
that when middle shafts are included
ends of the bones after hominids had that carnivore ravaging is density-me- in an analysis, the relative repre-
discarded the bones. Actualistic r-e- diated; middle shafts are more dense sentation of bones of different eco-
search showed that hyenas do remove than the ends of long bones and thus nomic utility also changes and, thus,
the ends of long bones from assem- survive better. I doubt that hyenas the resulting interpretations of behav-
blages that were processed in a manner make decisions about consuming long ior also ~ h a n g e . ~This
~ ~ sis true be-
typical of h ~ m i n i d sExperimental
.~ ta- bone ends based on their density. cause the magnitude of difference
between the density of the ends and
phonomy built on this earlier research Rather, they probably choose the ends
middle shafts of long bones is differ-
by more precisely documenting the ef- based on the presence of grease and
ent in different bones. For example,
fect of spotted hyenas on intrabone attached ligament mass. The density
the density of the middle shaft of the
survival.6 Estimates of the original of bones from a size one or two mam- tibia, a bone of fairly high utility, and
number of bones become more accu- mal is not of concern to spotted hye- that of the metatarsal, a bone of low
rate further from epiphyses or ends nas-they simply swallow the entire utility, is nearly the same (see Lyman,
(see Fig. 3). Unfortunately, as cine end of the bone with little preparatory Tables 7.6 and 7.7). However, the
moves farther from the ends of bone chewing. The positive relationship be- proximal end of the tibia is much less
specimens, their identifiability gener- tween long bone survival and bone dense than is the proximal end of the
ally decreases. density arises from the generally in- metatarsal. Thus, if we based our
Lyman discusses these issues and, verse relationship between density analysis of long bone abundance
extensively using my own data, argues and bone grease: that is, less dense tra- solely on ends, if tibias and metatar-
BOOK REVIEW Evolutionary Anthropology 71

sals were originally present in equal if there has been absolutely no den- worth the wait because they tell a
numbers, and if density-mediated de- sity-mediated destruction-that is, if story of limb-bone abundance as yet
struction had occurred, the metatar- carnivores or other destructive proc- unrecorded in other studies of Mous-
sals would appear t o be more esses have had no effect on the assem- terian assemblages.
abundant than tibias due the greater blage. If estimates based on middle
density of the metatarsal proximal shafts do not yield a greater abun- CONCLUSIONS
portion. If we then calculated the dance of long bones than ends, and if Taphonomy and zooarcheology are
abundance of long bones based on the analyst knows or suspects that the dynamic sister fields of investigation.
middle shafts, the picture would assemblage was subject to density- Lyman’s wonderful book captures ad-
change because we would expect, mediated destruction (for example, mirably the exciting progress we have
given the same conditions, that tibias the assemblage is a hyena den or is a made and, I hope, will lead to better
would be as abundant as metatarsals. human assemblage that has been rav- integration of disparate research. I
The data from the Mousterian site aged by hyenas), then either the meth- think that one of the books most im-
of Kobeh (western Iran), currently un- ods used for identification a n d portant effects will be to stimulate
der study in my laboratory, illustrate quantification were insensitive to the zooarcheologists to structure their
this point nicely Figure 3 shows the contribution of middle shafts or the methods and reporting of results in a
MNEs per bone portion for the meta- middle shafts were discarded at exca- way that will be more conducive to ta-
tarsals and tibias of bovids and cervids vation or curation (based on my obser- phonomic analysis.
size one and two, from those levels vations, this was a common practice We still have some serious deficien-
now completed. If we had restricted even up to the 1970s). cies in the field, including the few I
our study to the more identifiable ar- At the very least, zooarcheologists have covered here. My choices for dis-
ticular ends, metatarsals (MNE=7) should present middle-shaft data in a cussion were based in part on prob-
would be more abundant than tibias complete format (MNEs by bone por- lems I have attempted to address
(MNE=3). This pattern, following tion) and discuss how they identified through my own actualistic research.
Bl~menschine’s~~ criteria, could sug- the middle shafts (with or without However, I feel that each of these is-
gest that the hominids had late access conjoining). Lyman, in chapter 5, dis- sues is timely and looms large in
(scavenging) to the carcasses. But if cusses in detail the significance of zooarcheology as a whole. Obviously,
we include the middle shafts, the situ- bone conjoining studies, but neglects Lyman’s book stimulated my discus-
ation changes dramatically: tibias to mention their importance to shaft sion-ultimately, that is what excel-
(MNE=55)arenow more than twiceas identification. Many shaft fragments lent books are meant to accomplish. I
abundant as metatarsals (MNE=20). simply cannot be identified in the state believe that a great deal of discussion
This is more consistent with hominids in which they are discovered. How- and research will arise out of Lyman’s
having early access (hunting) to the ever, with patience and time, I have important achievement.
carcasses. The result is that found that many supposedly unidenti-
zooarcheologicalstudies that focus on fiable fragments become identifiable ACKNOWLEDGMENTS
long bone ends will tend to overesti- through conjoining. In the Kobeh as- This review benefited from the
mate the abundance of low-utility semblage, it is common for an identi- helpful comments of Robert Blumen-
long bones and the resulting interpre- fiable conjoined unit to be composed schine, Diane Gifford-Gonzalez, and
tations will suffer. Zooarcheological of eight or more fragments that were Yin Lam. The analysis of the Kobeh
studies that focus on long bone ends unidentifiable in isolation. material was assisted by Christopher
will also tend to overestimate the The study of middle shafts does cre- Norton, Godfrey Palaia, So0 Yeun
abundance of crania. Hyenas typically ate some problems. It is time-intensive Kim, Carol Frey, Anthony Rorke, and
chew through, but do not swallow, and highly insensitive to the academic Jennifer Teremy. Their hard work is
head parts. The result is that cranial calendar. Many of us who work in Af- gratefully acknowledged.
elements, which are reasonably easy rica, Europe, or Asia conduct our fau-
to identify, will appear to be more nal studies during the summer break. REFERENCES
abundant than long bones if the ana- A large assemblage of middle shafts 1 Gifford D (1981) Taphonomyand paleoecol-
lyst focuses only on long bone ends. simply cannot be completed in one, or ogy: A critical review of archaeology’s sister
even three summers. I and many discipline. In Schiffer MJ (ed) Advances in Ar-
Hyena assemblages or hyena-modi- chaeological Method and Theoq Volume 14, pp
fied assemblages may appear to have skilled technicians have been studying 365437. New York Academic Press.
high frequencies of crania relative to the Kobeh assemblage, which in- 2 Klein RG, Cruz-Uribe K (1984) The Analysis
of Animals Bones from Archaeological Sites.
long bones, but this is an artifact of the cludes approximately 20,000 speci- Chicago: University of Chicago Press.
method used to quantify the bones. mens, for more than three years, and 3 Gifford-Gonzalez D (1991) Bones are not
The following points can be pre- the work still is not complete. African enough: Analogues, knowledge, and interpre-
tive strategies in zooarcheology. J Anthropol
sented as axioms of zooarcheological faunas, because of their high number ArcheollO215-254.
methodology because they are based of like-sized species, are even more 4 Blumenschine RJ (1 987) Characteristics of
on solid actualistic research. The only complicated and time-consuming an early hominid scavenging niche. Curr An-
than are the species-poor Nearctic thropol28:383407.
situations in which estimates based on 5 Blumenschine RJ (1988) An experimental
middle shafts will not yield a greater faunas at Kobeh and areas such as model of the timing of hominid and carnivore
abundance of long bones than ends is Europe. However, I think the data are influence o n arcKaeoIogica1 bone assem-
72 EvolutionaryAnthropology BOOK REVIEW

blages. J Archeol Sci 15:483-502. 121-148. Boulder: Westview Press. positional destruction of bone in archaeologi-
6 Marean CW, Spencer LM (1 991) Impact of 13 Binford LR (1984) The Faunal Remains cal assemblages. J Archeol Sci 18:677-694.
carnivore ravaging on zooarcheological meas- from Klasies River Mouth. New York: Aca- 24 Klein RG (1975) Paleoanthropological im-
ures of element abundance. Am Antiquity demic Press. plications of the nonarchaeological bone as-
56:645458. 14 Klein RG (1989) The Human Career: Hu- semblage from Swartklip 1, south-western
7 Marean CW, Spencer LM, Blumenschine R J , man Biological and Cultural Origins. Chicago:
University of Chicago Press. Province, South Africa. Quatern Res 5:275-
Capaldo S (1992) Captive hyena bone choice
and destruction, the schlepp effect, and Oldu- 15 Binford LR ( I 978) Nunamiut Ethnoarchae- 288.
vai archaeofaunas. J Archeol Sci 19:101-121. ology. New York: Academic Press. 25 Bunn HT, Kroll EM (1986) Systematic
8 Blumenschine RJ,Marean CW (1993) A car- 16 Bunn HT, Bartram LE, Kroll EM (1988) butchery by Plio-Pleistocene hominids at
nivore’s view of archaeological bone assein- Variability in bone assemblage formation Olduvai Gorge, Tanzania. Curr Anthi-opol
blages. I n Hudson J (ed) From Bones to from HadLa hunting, scavenging, and carcass 27:43 1 4 5 2 .
Behavior, pp 273-301. Carbondale: Center for processing. J Anthropol Archeol 7412457.
Archaeological Investigation. 17 O’Connell JF, Hawkes K (1988) Hadza 26 Bartram LE, Kroll EM, Bunn HT (1991)
9 Stiner MC (1990) The use of mortality pat- hunting, butchering, and bone transport and Variability in camp structure and bone food
terns in archaeological studies of horninid their archaeological implications. J Anthropol refuse patterning at Kua San hunter-gatherer
predatory adaptations. J Anthropol Res 9:305- Res 44:113-1 6 1. camps. In Kroll EM, Price T D (eds) The Inter-
351. 18 OConnell JF, Hawkes K, Blurton-Jones N pretation ofArchaeologica1 Spatial Patterning,
10 Stiner MC (1991) An interspecific perspec- (1989) Reanalysis of large mammal body part pp 77-148. New York: Plenum Press.
tive on the emergence of the modern human transport among the Hadza. J Archeol Sci
17:30 1-3 16. 27 Blumenschine R J (1986)Carcass consump-
predatory niche. In Stiner MC (ed) Hunian tion sequences and the archaeological d,stinc-
Predators and Prey Mortality, p p 149-185. 19 Brain CK (1 981) The Hunters or the Hunted.
Boulder: Westview Press. Chicago: University of Chicago Press. tion of scavenging and hunting. J Hum Evol
20 Binford LR (198 1 ) Bones: Ancient Men and 15:639459.
1 1 Hudson J (1991) Nonselective small game
hunting strategies: An ethnoarchaeological Modem Myths. New York: Academic Press.
study of Aka pygmy sites. In Stiner MC (ed) 21 Lyman RL (1984) Bone density and differ- Curtis W. Mcrean
Human Predators and Prey Mortality, pp 105- ential survivorship of fossil classes. J Anthro- Department of Anthropology
120. Boulder: Westview Press. pol Archeol3:259-299. SUNY at Stony Brook
12 Blumenschine R (1991) Prey size and age 22 Lyman RL (1985) Bone frequencies: Differ-
ential transport, in situ destruction, and the Stony Broc k, NY
models of prehistoric hominid scavenging:
Test cases from the Serengeti. In Stiner MC MGUI. J Archeol Sci 12:221-236. curtis.marean.sunysb.edu
(ed) Human Predators and Prey Mortality. pp 23 Marean CW (1991) Measuring the postde- 0 1995 Wiley-Lks. Inc.

Meetings of Interest [email protected]. full cost of tuition will be $300 with

a reduced fee of $150 for students.
May 13-17, 1996 For further information and travel
March 27-3 1, 1996
Course in Forensic Anthropology plans contact John J. McIlwaine,
Society of Ethnobiology
Uniformed Services University of the Co-coordinator of Continuing and
19th Annual Conference
Santa Barbara, CA Health Sciences Professional Education, University
Bethesda. Maryland of Bradford, UK BD7 IDP,
Contact Jan Timbrook, Department
of Anthropology, Santa Barbara This five-day course surveys the tele 44-1274-385-482,
Museum of Natural History, 2559 basic principles of forensic fax 44- 1274-385-190.
Puesta del Sol Rd, Santa Barbara, anthropology and provides updates
CA 93105; 8051682-4711, ext 307; on new techniques in the field. A August 1 1-16, 1996
Fax 8051569-3170. new topic this year is the role of International PrimatologicalSociety /
forensic anthropology in mass American Society of Primatologists
April 9- 10, 1996 fatality situations. The course Joint Meeting
Paleoanthropology Society consists of a series of lectures Madison, Wisconsin
New Orleans, Louisiana covering topics in the field followed This joint meeting will be held at the
The Paleoanthropology meetings by laboratory sessions emphasizing University of Wisconsin, Madison,
will be held in conjunction with the hands-on analysis of skeletal and hosted by the Wisconsin
Society for American Archaeology at remains. For more information Regional Primate Center. Provisional
the New Orleans Marriott Hotel (555 write to the Center for Advanced registration costs are $150 for
Canal Street, New Orleans, LA Medical Education, Armed Forces regular members, $80 for students,
70140; tel: 5041581-1000,Fax: Institute of Pathology, Washington and $200 for non-members.
5041523-6755).Two full days of DC 20306-6000, tel(301) 295-7921 Registration includes opening and
papers are planned. or toll free at (800) 577-3749 or fax closing receptions, program, and
(301) 427-5001. abstract booklet. The conference
April 1 1 -13, 1996 includes oral presentations, posters,
American Association of Physical July 1-5, 1996 videos and film, symposia and
Anthropologists Course in Forensic Anthropology workshops. For more information,
65th Annual Conference Department of Archaeological Sciences contact Edith Chan, Wisconsin
Durham, NC University of Bradford, United Kingdom Regional Primate Center, 1223
For local arrangements, contact Dr: This five day course surveys the Capitol Court, Madison, WI 53715.
Matt Cartmill, Duke University basic principles of forensic Telephone: 608-263-3500;
Medical Center, Durham,NC 27710; anthropology and provides updates fax: 608-263-4031; email:
9191684-2971;Fax 9191684-8034;email: on new techniques in the field. The [email protected].

View publication stats