-
150 Lithic Technology, volume 22, no. 2
PLATFORM VARIABILITY AND FLAKE MORPHOLOGY:A COMPARISON OF
EXPERIMENTA L AND
ARCHAEOLOGICAL DATA ANDIMPLICATIONS FOR INTERPRETING
PREHISTORIC LITHIC TECHNOLOGICAL STRATEGIES
Harold L. Dibble
INTRODUCTION
At the heart of research on prehistoric lithicassemblages are
two fundamental questions. First,what were the processes by which
prehistoricflintknappers produced their implements? Andsecond, why
did they produce the forms they did?In the current literature a
great deal of attentionis paid to the factors that govern the
design ofvarious retouched types (e.g., Binford 1980;Torrence 1983,
1989: Bamforth 1986: Bleed1986; Bousman 1993; Andrefsky 1994;
Kuhn1994), their production (e.g., Sollberger 1977:Flenniken 1978;
Bradley and Sampson 1986;Whittaker 1994: Shott 1996). and their
mainte-nance (e.g., Gallagher 1977: Hayden 1977, 1979:Dibble 1984,
1987, 1995b; Barton 1988, 1990,1991; Neeley and Barton 1994).
Increasingly,however, attention is being paid to the technologyof
blank production: in fact, many believe thatthis is the most
potentially informative avenue ofresearch (e.g., Tixier et al.
1980; Inizan et al.1992; Sellet 1993).
Most technological studies, whether based onarchaeological
materials (e.g., Collins 1975: Marksand Volkman 1983; Volkman 1983:
Sullivan andRozen 1985: Baumler 1988: Cziesla et al. 1990)
orreplicative (flintknapping) experiments (e.g.,Crabtree 1966;
Bordes and Crabtree 1969; New-comer 1971; Mauldin and Amick 1989:
Ohnuma1995). have focused on core reduction sequences.While these
frequently tend to be heavily descrip-
. tive, i.e., simply documenting variability in reduc-tion
patterns (e.g., Boda 1994; Delagnes 1995),there is also a concern
for explaining technological
variability in terms of its adaptive significance,whether
relating it to raw material variability(Bietti and Grimaldi 1995;
Dibble et al. 1995).overall intensity of utilization (Baumler
1988:Dibble 1995b). or other factors (e.g., Munday1979: Marks 1983;
Henry 1989, 1995; Shea1995). The focus of this article is on the
relation-ship between certain platform variables and
flakemorphology, drawing on data from actual prehis-toric flake
assemblages and controlled experi-mental studies.
While it is undeniable that core reductionsequences represent a
significant aspect of lithic _technological variability, the
attention paid themhas resulted in a shift away from even
morefundamental processes of flaking. Some of theearliest works on
physical properties of flakedstone were those of Goodman ( 1944)
and Bourdier(1963), while Kerkhof and Mller-Beck (1969)were the
first to discuss concepts of Hertzianfracture and its application
to stone tools. Soonafter there was a virtual explosion of interest
infracture mechanics, with the works of Speth(1972,1974), Faulkner
(1972). Bonnichsen (1977).Cotterell and Kamminga
(1979,1986,1990),Cotterell et al. (1985), Tsirk (1974, 1979).
andBertouille ( 1989). However, the direct applicationof these
results to archaeological materials hasnot yet enjoyed a great deal
of success. There havealso been controlled experiments designed not
tofocus on fracture mechanics, but to documentrelationships between
particular independentvariables controlled by the flintknapper and
other,dependentvariables that are observable on flakes.While many
relationships have been found (Speth
Harold L. Dibble, Department of Anthropology, University of
Pennsylvania. Philadelphia, PA 19104 I
-
Dibble - Platform Variability and Flake Morphology 151
1975, 198 1; Dibble and Whittaker 198 1; Pelcin1986; Dibble and
Pelcin 1995; ), this approachhas not, in general, had a great
impact on lithicstudies, either.
One of the major problems in such experimentsis that the
experimental designs have not allowedfor the production of very
realistic-looking flakes.For example, in a series of experiments
performedby the present author (Dibble and Whittaker198 1: Dibble
and Pelcin 1995), flakes were re-moved along the sides of
plate-glass cores. Theflakes themselves more closely resembled
burinspalls than normal archaeological flakes or blades,and
platform width and flake width were alwaysconstant, equaling the
thickness of the plateglass. The design of the cores and the nature
of theflakes produced from them, plus the fact that somany
variables were held constant, resulted invery artificial products
that bore little resem-blance to most archaeological lithic
materials.Moreover, it is a fact that many variables --platform
morphology, core morphology, variousaspects of the delivery of
force when striking thecore, and the flaking qualities of the raw
materialitself -- interact during the production of a flake.Many of
these are invisible when analyzing reallithic artifacts. Legitimate
questions can there-fore be raised regarding the applicability of
theseresults to archaeological materials. What is clearlyneeded --
and this is the purpose of the presentpaper -- is to demonstrate
that the relationshipswhich became apparent in these controlled
ex-periments hold true for archaeological materialsand, moreover,
that they represent a significantaspect of prehistoric
technological variability.
This article deals primarily with the relation-ship between
certain platform variables -- plat-form thickness, platform width,
and exteriorplatform angle -- and flake morphology. A num-ber of
interesting findings come from the studiespresented below. First,
it is shown that theseplatform variables have a tremendous effect
onflake size and morphology, in spite of the fact thatso many other
potential factors are uncontrollablein dealing with archaeological
samples. Thus,platform variation and its effects on flake
mor-phology are fully applicable to archaeological in-vestigation.
Second, these platform variables areunder the direct control of the
flintknapper andpresumably are intentionally varied to
achieveparticular results. By focusing on these variableswhen
analyzing prehistoric assemblages, we can
add a new dimension to reconstructions of lithictechnology that
goes beyond core reduction pat-terns based on the sequencing and
direction offlake removals. Third, it is likely that some of
thealternative ways of preparing platform morphol-ogy have effects
that are both interpretable andexplainable in terms of current
models of techno-logical efficiency, resource economy, and
mobil-ity. Thus, attention to platform variation may con-tribute to
an overall understanding of technologi-cal variability in terms of
prehistoric adaptations.
MATERIALS AND METHODS
Two primary sources of data, derived from bothexperiments and
archaeological samples, are pre-sented here. Data from controlled
experimentsare drawn from a previously published study(Dibble and
Pelcin 1995) designed to assess thesignificance of harnmer mass and
velocity on flakemass. This sample contains a total of 177
com-plete flakes, produced from half-inch plate glasscores; the
flakes were removed along one edge bydropping steel ball bearings
on an adjacent edge.The two adjacent edges (one representing
theplatform surface and the other the exterior sur-face of the
core) were cut so as to produce exteriorplatform angles of 55, 65,
and 75 degrees. Angleof blow was held constant, but different sized
ballbearings were dropped from varying heights.
The archaeological data used in this paperrepresent many
different industries and technolo-gies. Chronologically they
include the Lower,Middle, and Upper Paleolithic: geographically
theyrepresent Africa, the Near East, and Europe; andtechnologically
they express varying emphases onLevallois, flake, biface, and blade
production (seeTable 1). Only complete, unretouched flakes areused
in this analysis and in all cases the total flakepopulation above 3
cm in maximum dimensionfrom a given assemblage was studied. While
thetotal sample size is high (over 12,000 completeflakes), the data
sets are not completely compa-rable: they were collected over a
period of morethan fifteen years in several different studies,
andfor some of them not all of the variables of interesthere were
recorded. Thus sample sizes for anyparticular analysis vary
considerably. In addi-tion, sample sizes for extreme values, when
bro-ken down by other variables, tend to get smallquickly. In the
analyses presented here, samplesizes of less than 5 cases were
eliminated.
-
Dibble - Platform Variability and Flake Morphology 153
For all of this material, including the experi-mental flakes,
measurements were taken in aconsistent manner even though the
actual record-ing was done by several different individuals.Length
is the distance from the point of percus-sion to the most distal
point on the flake. Widthis taken perpendicular to and at the
midpoint ofthe length axis: thickness, at the point where thelength
and width axes cross. Platform width is themeasurement along the
platform from one lateraledge to the other. Platform thickness is
measuredfrom the interior to the exterior surface of theplatform at
the point of percussion. It representsneither maximum platform
thickness nor theshortest distance from the point of percussion
tothe exterior surface. Bulb length was taken on theinterior
surface from the point of percussion to thebase of the bulb. All
measurements are expressedin mm.
Exterior platform angle, or the angle betweenthe platform
surface and exterior surface, is mea-sured with a goniometer
directly behind the plat-form, and expressed in degrees. On flakes
withcurved exterior surfaces, the measurement of thisangle can vary
depending on the point on theexterior surface at which the measure
was taken.For the sake of consistency, the angle used wasthat
formed by two lines -- one represented by theplatform thickness,
the other extending down theexterior face directly in line with the
axis ofpercussion to a distance equal to the platformthickness.
Weight was variously taken to thenearest 1 or 2 grams, depending on
the precisionof the scales being used. For the experimentalflakes,
platform width and flake width are con-stant (one-half inch) and
only three values ofexterior platform angles are represented.
A very conservative approach was taken withregard to all of
these measurements: if landmarkswere unclear, or if it looked as
though a portionwere missing, then the measurement was re-corded as
missing. Such problems were mostapparent for the platform
measurements, exteriorplatform angle, and bulb length. The
measure-ment of exterior platform angle is especially diffi-cult if
the platform surface was markedly curvedin the interior-exterior
plane just behind the pointof percussion or if the exterior surface
of the flakewas too irregular. Again, such cases were omittedfrom
consideration. On two separate occasionsthe level of inter-observer
reliability on this mea-sure was checked and the results found to
be
satisfactory. Both the definitions of these observa-tions and
the general conservatism in their re-cording should be kept in mind
when verifyingthese results with independent data.
Throughout this paper two words -- effect andassociation -- are
used seemingly interchangeablyto describe relationships
amongvariables, but it isimportant to recognize the difference in
theirmeanings. When the word effect is used, as inthe effect of
platform thickness on flake weight, itis because it is believed
that there is a trueindependent-dependent, i.e., causal,
relationshipbetween the two variables. Association, on theother
hand, means that two variables co-vary, butnot necessarily because
changes in one directlycause changes in the other. Many of the
analysesthat follow are based on showing relationshipsbetween
variables. By themselves, these relation-ships do not demonstrate
causal relationships,but only associations. In flintknapping,
however,many clear independent-dependent relationshipsdo exist,
given that the knapper can change suchfactors as how the blow is
delivered or the mor-phology of the core surface before the flake
isstruck. If other relevant variables are held con-stant, and
increases in platform thickness arecorrelated with increases in
flake weight, for ex-ample, then it is clear that the relation is
causal:platform thickness was determined prior to theformation of
the flake and, because other poten-tial variables were not factors
in producing vari-ability in weight, platform thickness is left as
thesole independent variable. Admittedly, when deal-ing with
archaeological assemblages there are agreat number of potentially
relevant variables thatcannot be controlled, so it is difficult to
assume acausal relationship between any two variables.This is why
controlled experiments, in whichmany variables are held constant,
are necessary,and it is one more reason why a comparisonbetween
experimentally-derived results and thosebased on archaeological
materials is crucial. None-theless, there are still many instances
in which ademonstrated relationship cannot be assumed tobe
causal.
Although there is a great deal of variation in thekinds of
assemblages represented here, the flakes,both experimental and
archaeological, representprimarily hard-hammer direct percussion
tech-niques in which the hammer struck a platformsurface. No claim
is made that the relationshipsfound here are directly applicable to
other tech-
-
.L
154 Lithic Technology, volume 22, no. 2
niques, such as bifacial thinning, bipolar, pres-sure flaking,
etc.
One final clarification is needed to facilitateinterpretation of
the analyses that follow. Invirtually every example except Figure
1, the indi-vidual data points shown on the graphs do notrepresent
individual cases, but rather averagevalues of the Y-axis variable
for all of the caseswithin a given interval of the X-axisvariable.
Thusthe probability statistics given in the figure leg-ends are
based on Analysis of Variance of thedifferences among the means and
not correlationsbased on individual flakes. This approach wastaken
because individual variability, especially inthe archaeological
data, is quite high; in fact, thereis often so much individual
variability that it tendsto obscure real relationships. This
variability is aresult of two things. First, it is a fact that
manyvariables contribute to flake morphology. If weare interested
in examining just the relationshipbetween platform thickness and
flake weight, forexample, it is impossible to choose flakes that
areconstant in every other measure. Thus individualflake
variability is high because of the varyingeffects of those other
variables, but the use ofaverages can help to isolate significant
trends.The second source of variability is measurementerror, a
result of either the difficulty of takingcertain measurements or of
problems in oper-ationalizing certain attributes. An example of
thesecond problem is platform area, which is theproduct of platform
thickness and platform width.This is a fine measure if the platform
surface isrectangular in shape, but it will vary from the truearea
if the platform surface is round, triangular, orvirtually any other
shape. To the extent that theseare random and not systematic
errors, individualvariability may be high, but the average values
willaccurately represent the central tendency.
RESULTS
Various controlled experiments have demon-strated that two
aspects of flake platforms --platform thickness and exterior
platform angle --have significant effects on flake size (Speth
1972,1974, 1981; Dibble and Whittaker 1981; Dibbleand Pelcin 1995).
This has been shown (Dibbleand Pelcin 1995) by plotting flake
weight againstplatform thickness for three different
exteriorplatform angles (see Figure 1). For each value ofexterior
platform angle, increasing the platform
thickness resulted in heavier flakes. Likewise, theslope of the
relationship between platform thick-ness and weight differed among
the three values ofexterior platform angle, such that higher
exteriorplatform angles resulted in increasingly heavyflakes for
the same value of platform thickness.
Another way of presenting these relationships,in the format that
will be used throughout theremainder of this paper, is shown in
Figures 2 and3. In Figure 2, for example, the average weights ofthe
experimental flakes with various platformthicknesses are plotted
and it is clear that in-creased platform thickness results in
larger flakes.In Figure 3, the relationship between
exteriorplatform angle and weight is shown broken downby different
values of platform thickness. Withineach interval, increasing the
exterior platformangle results in larger flakes. Note that the
largestaverage flake weights result from high values ofboth
platform thickness and exterior platformangle. In the
archaeological sample, increases inplatform thickness have a
similar effect on flakeweight, as shown in Figure 4.
Although the effect of platform width on flakeweight was not
addressed in the experiment be-cause of the design of the cores, it
is possible toexamine this relationship with the
archaeologicalsample. As can be seen in Figure 5, platform
widthexpresses a similar association with flake weight.However, the
relationship shown here is possiblymisleading: since platform
thickness tends to behighly correlated with platform width, it
could bethat the platform width-to-weight relationship isreally
only a reflection of the relationship betweenplatform thickness and
weight. In order to deter-mine the effect of platform width alone,
Figure 6displays the average weights of flakes by intervalsof
platform width within particular intervals ofplatform thickness. In
this way it can be seen that,for a particular interval of platform
thickness,increases in platform width result in larger
flakes.Given, then, that both platform thickness andplatform width
directly affect flake size, it shouldbe no surprise that platform
area (the product ofplatform thickness and platform width) also
showsa clear relationship with weight (Figure 7). Sinceplatform
area reflects the combined influence ofboth platform thickness and
platform width, all ofthe remaining analyses of archaeologicbe
based on it.
al data will
To demonstrate the effects of exterior platform
-
Dibble - Platform Variability and Flake Morphology 155
angle on weight, it is necessary to control for thecompeting
effects of platform area. Again, theclearest way to do this is to
plot exterior platformangle against weight by particular intervals
ofplatform area, which allows us to hold platformarea relatively
constant and thereby isolate theeffects of exterior platform angle
alone. When thisis done (Figure 8) the effects of exterior
platformangle on weight, for particular platform areas, areclear
and follow patterns similar to those seen inthe experimental data
(refer to Figure 3).
Since exterior platform angle and platform areaeach have a
direct relationship with overall weight,it would be expected that
increasing either or bothof these platform variables would also
increaseflake dimensions, i.e., length, width and thick-ness. This
is generally the case in both the experi-mental (Figure 9) and
archaeological (Figure 10)data. In comparing these two figures,
bear inmind that the experimental cores had a constantplatform
width, so that on these cores platformarea is directly and linearly
related to platformthickness (in other words, doubling the
platformthickness doubles the platform area). Thus theplatform
thickness of the experimental flakes isdirectly comparable to the
platform areas usedwith the archaeological data. Likewise, the
widthsof the experimental flakes was also constant, sotheir surface
areas are directly reflected by lengthalone.
Platform morphology alone is thus a significantfactor affecting
flake size, whether size is ex-pressed in terms of weight or
dimensions. Obvi-ously, flake weight is a direct function of the
threedimensions, but the importance of separatingthese two aspects
of size (weight versus dimen-sions) will be addressed in more depth
later in thepaper. For now, it is clear that a flintknapper hastwo
options for increasing overall flake size: pre-pare either higher
exterior platform angles orbigger platform areas. While these are
not mutu-ally exclusive options over a range of differentflake
sizes, to produce a flake of a given size thereis the possibility
of having a high platform areaand low exterior platform angle, or
conversely, alow platform area and high exterior platform
angle.Thus, using weight as a reflection of size, when welook at
the data for any single weight, these twovariables should be
inversely related to eachother. As shown in Figure 11, this is the
case forboth the experimental (uppergraph) and archaeo-logical
(lower graph) data. Thus, adjusting plat-
form area or exterior platform angle representstwo alternatives
open to the flintknapper.
It should be clear that, in every case in whichdirect
comparisons can be made, the archaeologi-cal data show the same
relationships betweenplatform morphology and flake size that are
seenwith flakes produced under controlled experi-ments. This is
important for a couple of reasons.First, it helps to validate the
results obtained fromsuch experiments. In spite of their
artificialconditions, they are useful for isolating and
objec-tifying relationships that are relevant to archaeo-logical
materials. Some of these relationships aredifficult to observe in
normal flintknapping simplybecause so many variables are operating
simulta-neously. Second, looking at it from the oppositepoint of
view, the successful extension of theexperimental results strongly
suggests that a largenumber of potentially significant
independentvariables (hammer type, angle of blow, etc.), whichwere
controlled in the experimental situation,exert only minor, if any,
effects on these relation-ships (see also Pelcin 1996).
Other factors can be examined in the archaeo-logical data to see
what effects they may have, ifany. Three types of comparisons that
reflectaspects of overall core reduction technologies arepresented
here: scar morphology, technologicalcategory, and industry. In
Paleolithic research(e.g., Dibble and Bar-Yosef 1995) attention
iscurrently focused on these kinds of variables asrepresenting a
major source of inter-assemblagevariability. Thus it is important
to see whether ornot different core reduction technologies affect
therelationships under investigation here.
In order to save space, the comparisons pre-sented here are
based on only two of the relation-ships shown above: platform
area-to-flake weight,and exterior platform angle to the ratio of
surfacearea to platform area. The first (Figure 12) is theeffect
that scar morphology has on these relation-ships, which appears to
be minimal. The second(Figure 13) examines the role that basic
technol-ogy plays, as recorded on flakes that are clearlydiagnostic
of one or another of three major tech-nological categories --
Levallois, blade, and nor-mal (or non-diagnostic). Admittedly,
these tech-nological classes are very superficial and
probablyencompass a tremendous amount of variability.Nonetheless,
they all show the same general rela-tionships.
-
l .
156 Lithic Technology, volume 22, no. 2
The third comparison (Figure 14) is brokendown by industry. Here
the number of potentialfactors that may affect the relationships in
ques-tion is quite high. Raw material is one of the moreimportant
of these. While most of these industriesare made on local flint,
the Nubian Khormusanpeople utilized ferracrete sandstone, while
theflints used at the other sites also differed in termsof size and
abundance. However, good data on theraw materials used in these
assemblages are notavailable in a way that could be
systematicallyincorporated into the present analysis. There arealso
clear technological differences among theseindustries, which result
in different core mor-phologies. The Khormusan, Levantine
Moust-erian, and Biache are all considered Levalloisindustries,
though the specific reduction sequencesvary considerably (cf. Boda
1988: Meignen andBar-Yosef 1988, 1991, 1992; Dibble 1995a;
Sellet1995). The Levantine Upper Paleolithic sample iscomposed of
both Ahmarian and LevantineAurignacian, which have varying emphases
onblade production. Level F4 from Pech de 1Az isMousterian of
Acheulian tradition and includessome biface manufacture, while
Combe-Capellehas relatively little Levallois production and
nobifaces. Although the various industries aresomewhat separate
from each other, which pre-sumably reflects the effects of these
uncontrolledvariables, the basic relationships are still appar-ent
within any one of the assemblages.
What these figures suggest is that, while corereduction
technologies differ among Paleolithicassemblages, they do so
independently of thekinds of relationships examined here. Thus
plat-form morphology and its effects on flake morphol-ogy represent
a completely separate dimension oflithic technology and flake
assemblage variability.
technological importance for the production offlakes. The
aspects of the platform that wereaddressed here (platform
thickness, platformwidth, and exterior platform angle) are all
underthe direct control of the flintknapper, and they aretrue
independent variables in that they directlyaffect flake morphology.
This is not to say thatthey represent the only factors affecting
flakemorphology, and they are certainly not the onlyimportant
factors giving rise to lithic assemblagevariability. But they do
have major effects andundoubtedly reflect strategies employed by
theflinknapper to produce different results. Theyrepresent aspects
of lithic technology that arerelevant to behavior and should not be
ignored inreconstructing prehistoric technologies.
It should be emphasized again that these re-sults have relevance
only for flakes produced bydirect percussion on a platform surface.
Theextent to which they are applicable to flakesproduced by
pressure is not known, though it islikely that the relationships
would remain essen-tially the same. The results obtained from
othertechniques, such as bifacial flaking (in which theexterior
edge of the platform is struck) or bipolarflaking, are probably not
directly comparable. Italso bears repeating that, as clear as these
rela-tionships are, there is a great deal of variabilityamong
individual flakes. It is quite easy to findlarge flakes with small
platforms and small flakeswith large platforms. Many factors
interact incomplex ways to give rise to flake morphology andresults
are not always completely predictable. Onthe other hand, even with
only a moderate amountof experience, a flintknapper can achieve a
signifi-cant degree of control over flake variability. Byisolating
the contribution that platform morphol-ogy makes on flake
variability, our understandingof different flaking strategies is
enhanced, even ifthat contribution is less than 100 percent.
DISCUSSION AND CONCLUSIONSBeyond the demonstration of the
relationships
between platform morphology and flake morphol-ogy, what is
interesting is that prehistoric knappersadopted different
strategies with respect to howthey set up platform morphology. For
example,among the assemblages included here, there is ahigh degree
of variability in terms of platform areaand exterior platform angle
(Figure 15). Someassemblages -- for example, the Levantine
Mous-
Platform area alone has been utilized for sometime as a means to
control for original flake size inassessing the degree of reduction
that took placeon a flake tool from retouching (Dibble 1987,1995b).
The results presented here support suchan approach, though they
also suggest that exte-rior platform angle should be taken into
accountas well, in assessing flake size. The implicationsof these
results go beyond attempts to reconstructor control for original
flake dimensions. Theysuggest that platform characteristics are of
major
terian from the site of D40 -- seem to maximizeboth exterior
platform angle and platform area.Others appear to minimize platform
area but
-
Dibble - Platform Variability and Flake Morphology 157
exhibit relatively high exterior platform angles, orvice versa.
Although the choice of adopting onestrategy over another may simply
reflect differenttechnological traditions, it may also be related
toa number of other factors.
As shown above, increasing either platformarea or exterior
platform angle increases overallflake size. Industries that have,
on average, highvalues for both probably reflect an attempt
tomaximize the size of the resulting blanks. Thismay reflect
conditions in which raw material waslarge and plentiful. But
differences also resultfrom increasing platform area, on the one
hand,and exterior platform angle, on the other.. Thesemay reflect
different flintknapping strategies thatmay also be related to
either raw material or otherconsiderations.
The choice of one or the other of these ap-proaches is related
to differences in flake mor-phology, especially with respect to
flake dimen-sions relative to platform size. If flakes are
madebigger by increasing the size of their platforms,then there is
a general decrease in flake dimen-sions relative to platform size
(Figures 16 and 17).Adjusting the exterior platform angle shows
theopposite association: flake dimensions in relationto platform
size increase with higher exteriorplatform angles. For the
archaeological sampleonly the ratio representing flake area to
platformarea is shown, but there is a similar effect for bothlength
and width individually. Thus, increasingthe exterior platform angle
results in larger flakesrelative to their platform areas.
This difference, which is illustrated schemati-cally in Figure
18, has a couple of importantimplications. First, there is a
consideration of coremaintenance. By producing larger flakes
withlarger platforms, more of the edge is removed asthe striking
platform for a given flake area. If theexterior platform angles are
increased instead,then the striking platform of a core is
conservedlonger, thereby extending the usefulness of thecore in
relation to the size of the flakes taken fromit.
Second, platform area represents, for mostpurposes, wasted
material on a flake blank: itprovides no cutting edge, and is
rarely retouched,and its overall size may constrain hafting.
Fur-thermore, it can be shown in the archaeologicaldata that bulb
length is also associated with
platform area (Figure 19), though it is likely thatplatform
width contributes more than platformthickness to bulb length. Given
that the bulb isitself a function of platform size and it, too,
repre-sents a less desirable portion of the total flakeweight, it
would not always be advantageous fora flintknapper to follow a
strategy of increasingflake size by increasing platform size. By
produc-ing flakes with higher exterior platform angles,however,
s/he can increase flake dimensions whilesimultaneously holding down
the size of both theplatform and bulb. This would help maintain
corestriking platforms and, at themore efficient flake blanks.
same time, produce
On the other hand, employing higher exteriorplatform angles has
disadvantages. As shown bySpeth (1981) and Dibble and Pelcin
(1995), therange of both force and angle of blow that willproduce a
flake is considerably less with higherexterior platform angles than
with lower ones.Therefore, it is generally easier to produce
flakeswith lower exterior platform angles. If raw mate-rial is
abundant, then core maintenance may notbe a significant
consideration and smaller exteriorplatform angles may be employed.
If it is scarce,then increasing flake exterior platform angleswould
help both to economize cores and to pro-duce more efficient flake
blanks, though the de-gree of control necessary in flaking is
increased.With these considerations in mind, it is interest-ing to
note that both D40 and Combe-Capelle areprimary exploitation sites
in the immediate vicin-ity of raw material (Munday 1976; Marks,
per-sonal communication; Dibble and Lenoir 1995).and both are among
those sites with the highestplatform areas. There may also be
considerationsinvolved with the maintenance of high platformangles
through faceting or other methods of coreplatform rejuvenation.
Certain aspects of platform variability de-scribed here also
have implications for recentdiscussions of flake size and group
mobility, inwhich it has been argued (Kuhn 1994, 1996; cf.Morrow
1996), based fundamentally on surfacearea to weight ratios, that
smaller flakes are moreefficient in terms of transport costs. It is
true thatallometry alone would predict that smaller flakeshave
greater surface area to weight, since surfacearea increases as the
square, while weight in-creases as the cube. If all dimensions
increaseequally, then larger flakes, even though they couldsustain
more resharpening because of their abso-
-
. .
158 Lithic Technology, volume 22, no. 2
lute size and thus have higher potential use-lives,would
necessarily have smaller surface area toweight ratios and thus
would represent less effi-cient objects for transport. It has been
repeatedlydemonstrated, however, that larger blanks
wereconsistently selected for tools and transportedduring the
Paleolithic, at least (Geneste 1985;Dibble 1988, 1995a, Dibble and
Holdaway 1993;Meignen 1993; Dibble et al. 1995;). If larger
flakesare less efficient, then why are they consistentlythe ones
being selected and transported?
Assuming that flakes will be reduced until theyreach some
minimum size (Dibble 1995b, Kuhn1994), larger flakes do have an
advantage in thata smaller proportion of their overall size is
simplywaste. As pointed out by Kuhn (1996), however,the real issue
is not overall size, but shape: forflakes with the same weight,
thinner flakes withlarger surface areas are more efficient (in
terms ofresharpening potential) than thicker flakes thathave less
surface area. Such a change overcomesthe predicted allometric
relationship, and thusrepresents a possible technological strategy
thatcould be advantageous under certain circum-stances. In fact, it
has been shown at Combe-Capelle (Roth et al. n.d.) that transported
flakes(those imported into the site) do exhibit highersurface area
to weight ratios than those that weremanufactured from immediately
available mate-rial and left behind. So it seems likely that at
leastsome Paleolithic peoples recognized the advan-tage of such
flakes when they selected them fortransport.
The archaeological data presented here alsosuggest that knappers
influence such differencesin flake shape by changing their platform
mor-phology. In these data, varying ratios of platformwidth to
platform thickness are associated withdifferent ratios of flake
surface area to thickness.When both exterior platform angle and
platformarea are controlled, the ratio of platform width toplatform
thickness does not have a significantassociation with overall flake
weight (Figure 20,upper graph). However, producing larger plat-form
widths relative to platform thicknesses clearlychanges the
distribution of weight, expressed byflakes with increasingly higher
ratios of surfacearea to thickness (Figure 20, lower graph).
Thus,one way to increase a flakes surface area tothickness, and
therefore make it more efficient,would be to increase platform
width relative toplatform thickness. Moreover, even when we
look
at assemblage averages (Figure 2 l), the relation-ship between
platform width to thickness andflake surface area to thickness is
clear. Thedistribution shown here may indicate that some ofthese
industries were intentionally providing moreefficient flakes for
transport.
These are just two examples to show thatunderstanding platform
morphology and its ef-fects on flake morphology may relate to an
overallunderstanding of the adaptive significance of
lithictechnological variability. Further research isneeded in this
area, and it would be premature toclaim that flake platform
variability is more impor-tant than other aspects of lithic
technology. Onthe other hand, it cannot be denied that it
repre-sents a significant factor, and one with potentialfor
interpreting assemblages in terms of pastbehavior and adaptation.
What must be kept inmind is that platform variability reflects
actionsthat a flintknapper takes prior to the removal of asingle
flake, presumably to achieve certain de-sired effects.
Understanding what those effectsare should help in recognizing
particular strate-gies employed by prehistoric flintknappers.
Knowing the effects ofvarious platform charac-teristics on flake
morphology, however, is onlyone aspect of the problem. There is
also a need tounderstand and recognize different options
forcontrolling them. There are probably many differ-ent ways to
change platform angles, for example;and likewise, platform shape
can vary as a resultof many different actions. Moreover, there is
anobvious need to understand more of the effects ofcore surface
morphology on flake variation, espe-cially lateral and longitudinal
convexity. This ismore than just a question of scar patterns,
sincesimilar core morphologies can be produced indifferent ways.
.
As stated at the beginning of this paper, pat-terns of core
reduction sequences represent animportant aspect of lithic
technology, and theidentification of particular reduction patterns
is ajustifiable goal in lithic research. It is hoped,however, that
a case has been made here thatunderstanding even more fundamental
aspects offlintknapping also can provide insights into howand why
lithic assemblages vary. By adding thisdimension to our analytical
arsenal, our under-standing of prehistoric lithic technologies
willimprove.
-
Dibble - Platform Variability and Flake Morphology 159
ACKNOWLEDGMENTS
For access to collections I would like to thankArthur Jelinek,
Anthony Marks, Denise deSonneville-Bordes, Alain Tuffreau, and the
JerseyMuseum: for help in data collection, and forvarious
discussions about flake variability, thanksgo to Joshua Beeman,
Philip Chase, SimonHoldaway, Shannon McPherron, April Nowell,Andrew
Pelcin, Barbara Roth, Helen Sanders-Grey, and reviewers of this
manuscript; for fund-ing I thank the National Science Foundation
(Grant#s BNS 8804379 and DBS 92-20828), the De-partment of
Anthropology, Research Foundation,the University Museum of the
University of Penn-sylvania, and private donors.
Andrefsky, W.1994 Raw Material Availability and the
Organiza-
tion of Technology. American Antiquity59:2 l-35.
Bamforth, D.1986 Technological Efficiency and Tool Curation.
American Antiquity 5 1:38-50.Barton, C. M.
1988 Lithic Variability and Middle PaleolithicBehavior. British
Archaeological Reports,International Series no.408. Oxford
1990 Stone Tools and Paleolithic Settlement in theIberian
Peninsula. Proceedings of the Pre-historic Society 56: 15-32.
199 1 Retouched Tools: Fact or Fiction? Paradigmsfor
Interpreting Paleolithic Chipped Stone. InPerspectives in
Prehistory, edited by G. A.Clark, pp. 143- 163. University of
Pennsylva-nia Press, Philadelphia.
Baumler, M.1988 Core Reduction, Flake Production, and the
Middle Paleolithic Industry of Zobiste (Yugo-slavia). In Upper
Pleistocene Prehistory ofWestern Eurasia, edited by H. L. Dibble
andA. Montet-White, pp. 255-274. UniversityMuseum, University of
Pennsylvania, Phila-delphia.
Bertouille, H.1989
Bietti, A1995
Theories physiques et mathkmatiques dela taille des out&
pr&historiques. CNRS,Cahiers du Quaternaire, no. 15. Paris.and
S. GrimaldiLevallois Debitage in Central Italy:
TechnicalAchievements and Raw Material Procurement.In The
Definition and Interpretation ofLeuaZZois Variability, edited by H.
L. Dibbleand 0. Bar-Yosef, pp. 125-142. PrehistoryPress,
Madison.
Binford, L.1980 Willow Smoke and Dogs Tails: Hunter-Gath-
erer Settlement Systems and ArchaeologicalSite Formation.
American Antiquity 45:4-20.
Bleed, P.1986 The Optimal Design of Hunting Weapons:
Maintainability or Reliability. American An-tiquity
51:737-747.
Boeda, E.1988 Analyse technologique du debitage du niveau
IIA. In Le Gisement Pal&olithique Moyen
deBiacheSaint-Vaast(Pasxk-CaZais), Volume1, edited by A. Tuffreau
and J. Sommi, pp.185-2 14. Memoir-es de la Societe
PrehistoriqueFranc&e, Volume 2 1. Paris.
Boeda, E.1994 Le concept Levallois: Variabilitk des
Mdthodes. CNRS, Paris.Bonnichsen, R.
1977 Models for Discovering Cultural I@iorma-tion from Stone
Too&. National Museum ofCanada, Archaeological Survey of
Canada,Paper no. 60. Ottawa.
Bordes, F.1955 La stratigraphie de la grotte de Combe-Grenal
(Dordogne). Note preliminaire. BuZZetin dela
Soctit@r&historiqueI+a~aise52:426-429.
1976 Le gisement de Pech de 1AzC IV. Notepreliminaire. Bulletin
de la Soci&t&Prkhistotique I+anCaise 72:293-308.
Bordes, F. and D. Crabtree1969 The Corbiac Blade Technique and
Other Ex-
periments. Tebiwa 12: l-2 1.Bourdier, F.
1963 Sur la genese et la morphologie de leclatprehistorique.
Comptes Rendus deIAcadkmie des Sciences (Paris) 257:3975-3978.
Bousman, C.1993 Hunter-Gatherer Adaptations, Economic Risk
and Tool Design. Lithic Technology 1:59-86.
Bradley, B., and C. G. Sampson1986 Analysis by Replication of
Iwo Acheulian
Artefact Assemblages from Caddington, En-gland. In
StoneAgePrehistory, edited by G.Bailey and P. Callow, pp. 29-46.
CambridgeUniversity Press, Cambridge.
Callow, P., and J. Cornford (editors)1986 La GrottedeSt.
Brelade, 1961=1978Exca-
vations by C.B.M. McBumey. Geo Books,Norwich.
Collins, M.1975 Lithic Technology as a Means of processual
inference. In Lithic Technology: Makingand Using Stone Tools,
edited by E. Swanson,pp. 15-34. Mouton, The Hague.
-
160 Cithic Technology, volume 22, no. 2
Cotterell B., and J. Kamminga1979 The Mechanics of Flaking. In
Lithic Use-
WearAnalysis, edited by B. Hayden, pp. 97-1 12. Academic Press,
New York.
1986 Finials on Stone Flakes. Journal ofArchaeo-logical Science
13:451-461.
1990 Mechanics of Pre-Industrial Technology.Cambridge University
Press, Cambridge.
Cotterell, B., J. Kamminga and F. P. Dickson
1995a
1995b
Biache-Saint-Vaast, Level IIa: a Comparisonof Approaches. In The
Definition and Inter-preta tion of LeuaZZois Variability, editedby
H. L. Dibble and 0. Bar-Yosef, pp. 93- 116.Prehistory Press,
Madison.Middle Paleolithic Scraper Reduction: Back-ground,
Clarification, and Review of Evidenceto Data. Journal of
Archaeological Methodand Theory 2:299-368.
1985 The Essential Mechanics of Conchoidal Flak- Dibble, H. L.,
and 0. Bar-Yosef (editors)ing. Internationat JoumaZ of Fracture
1995 The Definition and Interpretation of20:205-22 1. Leualtois
VariabUity. Prehistory Press, Madi-
Crabtree, D. son.1966
Crew, H1976
A Stoneworkers Approach to Analyzing andReplicating the
Lindenmeier Folsom. Tebiwa9:3-39.L.The Mousterian Site of Rosh Ein
Mor. InPrehistory and Paleoenuironments in theCentral Negeu,
Israel, VolumeI: theAvdat/Aqeu Area, Part 1, edited by A. E. Marks,
pp.75- 112. Southern Methodist University Press,Dallas.
Cziesla, E., S. Eickhoff, N. Arts, and D. Winter (editors)1990
TheBig Puzzle: International Symposium
on Refitting Stone Artefacts. Holos Verlag,Bonn.
Debenath, A. and A. J. Jelinek1987 Gardes-Le Pontaroux, La
Quina. Notice sur la
campagne de fouilles 1986. Bulletin Liaisonet Information,
Associat ion desArch&oZogues de Poitou-Charentes 16: 14-15.
Delagnes, A.1995 Variability within Uniformity: Three Levels
of
Variability within the Levallois System. InThe Defanition and
Interpretation ofLevaZZois Variability, edited by H. L. Dibbleand
0. Bar-Yosef, pp. 201-2 11. PrehistoryPress, Madison.
Dibble, H. L.1981 Technological Strategies of Stone Tool
Production at Tabun Caue (Israel). Depart-ment of Anthropology,
University of Arizona,Tucson.
1984 Interpreting Typological Variation of MiddlePaleolithic
Scrapers: Function, Style, or Se-quence of Reduction? Journal of
Field Ar-chaeology 11:43 l-436.
1987 The Interpretation of Middle PaleolithicScraper Morphology.
American Antiquity52:109-117.
1988 Typological Aspects of Reduction and Inten-sity of
Utilization of Lithic Resources in theFrench Mousterian. In Upper
PleistocenePrehistory of Western Eurasia, edited by H.L. Dibble and
A. Montet-White, pp. 181-187.University Museum Symposium Series,
Vol I.Philadelphia.
Dibble, H. L., and S. Holdaway1993 The Middle Paleolithic
ofwarwasi Rockshelter.
In ThePateoZithicI+ehistoryof theZagrw,edited by D. Olszewski
and H. L. Dibble, pp.75-99. University Museum Press,
Philadel-phia.
Dibble, H. L., and M. Lenoir (editors)1995 The Middle
Paleotithic Site of Corn&-
Capelle Bas (France). The University Mu-seum Press,
Philadelphia.
Dibble, H. L., and A. Pelcin1995 The Effect of Hammer Mass and
Velocity on
Flake Mass. Journal of Archaeological Sci-ence 22:429-439.
Dibble, H. L., B. Roth, and M. Lenoir1995 The Use of Raw
Materials at Combe-Capelle
Bas. In The Midcite Paleolithic Site ofCornbe-Capelle Bas
(France), edited by H. L.Dibble and M. Lenoir, pp. 259-287. The
Uni-versity Museum Press, Philadelphia.
Dibble, H. L., and J. Whittaker1981 New Experimental Evidence on
the Relation
between Percussion Flaking and Flake Varia-tion. Journal of
Archaeologicat Science6:283-296.
Faulkner, A.1972 Mechanical Principks of Flintworking.
PhD dissertation, Department of Anthropol-ogy, Washington State
University. UniversityMicrofilms, Ann Arbor.
Ferring, C. R.1977 The Late Upper Paleolithic Site of Ein
Aqev
East. InPrehistory andPaZeoenuirwnmentsin the Central Negeu,
Israel. VoZumeII: TheAwiat/AqevArea,Part 2 and theHarHa*edited by
A. E. Marks, pp. 81- 118. SouthernMethodist University Press,
Dallas.
Flenniken, J.1978 A Reevaluation of the Lindenmeier Folsom:
a
Replication Experiment in Lithic Technology.American Antiquity
43:473-480.
Gallagher, J.1977 Contemporary Stone Tools in Ethiopia:
Impli-
cations for Archaeology. Journal of FieldArchaeology 4:407-4
14.
Geneste, J.-M.1985 Analyse Zithique dindustries mous-
-
Dibble - Platform Variability and Flake Morphology 161
t&iennes du Pdrigord: une approchetechnologique du
comportement desgroupes humains au PaZeSoZithique Moyen.Universitc
de Bordeaux I.
Goodman, M. E.1944 The Physical Properties of Stone Tools
Mate-
rials. American Antiquity 9:4 15-433.Hayden, B.
1977 Stone Tool Functions in the Western Desert.In Stone Took as
Cultural Markers :Change, Evolution and Complexity, editedby R. V.
S. Wright, pp. 178 188. AustralianInstitute of Aboriginal Studies,
Canberra.
1979 PateoZithicReflectionst Lithic Technologyand Ethnographic
Excavations amongAustralian Aborigines. Australian Instituteof
Aboriginal Studies, Canberra.
Henry, D. 0.1989
1995
Correlations between Reduction Strategiesand Settlement
Patterns. In Alternative Ap-proaches to Lithic Analysis, edited by
D. 0.Henry and G. Odell, pp. 139- 156. Archaeo-logical Papers of
the American Anthropologi-cal Association, no. 1. Washington.The
Influence of Mobility Levels on LevalloisPoint Production, Late
Levantine Mousterian,Southern Jordan. In The Definition
andInterpretation of Levallois Variability, ed-ited by H. L. Dibble
and 0. Bar-Yosef, pp. 185-200. Prehistory Press, Madison.
Inizan, M.-L., H. Roche, and J. Tixier1992 Technology of Knapped
Stone. Prkhistoire
de ZaPierre TaiZZGe, Tome 3. CREP, Meudon.Jelinek, A. J.
1982
Kerkhof,1969
Kuhn, S1994
1996
Tabun Cave and Paleolithic Man in the Le-vant. Science 2 16:
1369- 1375.F., and H.-J. Muller-Beck.Zur bruchmechanischen Deutung
derSchlagmarken an Steinger&ten. Glas tech-nische Berichte
42:439-448.L.A Formal Approach to the Design and Assem-bly of
Mobile Toolkits. American Antiquity59:426-442.The Trouble with Ham
Steaks: a Reply toMorrow. American Antiquity 6 1:59 l-596.
Marks, A. E.1968
1976
1977
The Khormusan: an Upper Pleistocene Indus-try in Sudanese Nubia.
In The Prehistory ofNubia 1, edited by F. Wendorf, pp. 3 15-39
1.Southern Methodist University Press, Dallas.Ein Aqev: a Late
Levantine Upper PaleolithicSite in the Nahal Aqev. In Prehistory
andPaleoenvironments in the Central Negev,Israel. Volume I: the
Avcfat/Aqev Area,Part 1, edited by A. E. Marks, pp.
227-292.Southern Methodist University Press, Dallas.The Upper
Paleolithic Sites of Boker Tachtitand Boker: a Preliminary Report.
In Prehis-tory and Paleoenvironments in the Cen-
tral Negev, Israel. Volume II: The Avdat/AqevArea, Part 2 and
the HarHar& editedby A. E. Marks, pp. 61-80. Department
ofAnthropology, Southern Methodist Univer-sity, Dallas.
1983 The Middle to Upper Paleolithic Transition inthe Levant. In
Advances in World Archaeol-ogy, edited by F. Wendorf and A. Close,
pp.5 l-98. Academic Press, New York.
Marks, A. E., and P. Volkman1983 Changing Core Reduction
Strategies : a Tech-
nological Shift from the Middle to the UpperPaleolithic in the
Southern Levant. In TheMousterian Legacy, edited by E. Trinkaus,pp.
13-34. British Archaeological Reports,International Series 164.
Oxford.
Mauldin, R., and D. Amick1989 Investigating Patterning in
Debitage from
Experimental Bifacial Core Reduction. In a-pediments in Lithic
Technology, edited byD. Amick and R. Mauldin, pp. 67-88.
BritishArchaeological Reports, International SeriesS528.
Oxford.1Meignen, Ue
1993
1988
1991
1992
Morrow,1996
Munday,1976
1979
LAbri des Canalettes. Un h a b i t a tmoust&ien sur ks
Grands Causses (Nant,Aveyron). Fouilles 1980-1986. Centre Na-tional
de la Recherche Scientifique, Paris.Meignen, L., and 0.
Bar-YosefVariabilite technologique au Proche-Orient:lexample de
Kebara. In LHomme deNeandertat, Vol. 4: La Technique, editedby L.
Binford and J.-P. Rigaud, pp. 8 l-96.Liege.Les outillages lithiques
moust&iens de Kebara(Fouilles 1982-1985). In Le
SqueletteMoustirien de K&ara 2, edited by 0. Bar-Yosef and B.
Vandermeersch, pp. 49-76. Edi-tions du CNRS, Cahiers de
Paleontologie,Paris.Middle Paleolithic Variability in Kebara
Cave,Mount Carmel, Israel. In The Evolution andDispersal of Modern
Humans in Asia, ed-ited byT. Akazawa, K. AokiandT. Kimura, pp.129-
148. Hokusen-Sha, Tokyo.
I. A.I
Bigger Is Better: Comments on Kuhns For-mal Approach to Mobile
Tool Kits. AmericanAntiquity 61:581-590.F.Intersite Variability in
the Mousterian Occu-pation of the Avdat/Aqev Area. In Prehistoryand
Pakoenvironments in the CentralNegev, Israel. Volume II: The
Avdat/AqevArea, Part 2 and the Har Ha%, edited by A.E. Marks, pp.
113-140. Department of An-thropology, Southern Methodist
University,Dallas.Levantine Mousterian TechnologicalVariabil-ity: a
Perspective from the Negev. Pal&orient5:87- 104.
-
162 Lithic Technology, volume 22, no. 2
Newcomer, M.1971 Some Quantitative Experiments in Handaxe
Manufacture. World A~haeology 3:85-94.Neeley, M., and C.M.
Barton
1994 A New Approach to Interpreting Late Pleis-tocene Microlith
Industries in Southwest Asia.Antiquity 68:275-288.
Ohnuma, K.1995 Analysis of Debitage Pieces from Experimen-
tally Reduced Classical Levallois and Dis-coidal Cores. In The
Definition and Inter-pretation of Levallois Variabitity, editedby
H.L. Dibble and 0. BarYosef, pp. 257-266.Prehistory Press,
Madison.
Pelcin, Andrew1996
Sellet, F.1993
1995
Shea, J.1995
ControZkdExperiments in theProductionof Flake Attributes. Ph.D.
Dissertation,Department of Anthropology, University
ofPennsylvania.
Cha-l^ne Operatoire: the Concept and Its Appli-cations. Lithic
Technology 1: 106- 112.Levallois or Not Levallois: Does It Really
Mat-ter? Learning from an African Case. In TheDefinition and
Interpretation of LevaZZoisVariability, edited by H.L. Dibble and
0.Bar-Yosef, pp. 25-39. Prehistory Press, Madi-son.
Behavioral Factors Affecting the Productionof Levallois Points
in the Levantine Moust-erian. In TheDefznition and Interpretationof
Levallois Variability, edited by H.L.
pact. American Antiquity 40:203-207.198 1 The Role of Platform
Angle and Core Size in
Hard-Hammer Percussion Flaking. LithicTechnology 10: 16-2 1.
Sullivan, A., and K. Rozen1985 Debitage Analysis and
Archaeological Inter-
pretation. AmericanAntiquity 50:755-779.Tixier, J., M.-L.
Inizan, and H. Roche
1 9 8 0 Prkhistoire de la Pierre tailthe 1:Terminotogie et
technologie. Cercle deRecherches et dEtudes Prehistoriques,Valbonne
, France.
Torrence, R.1983 Time Budgeting and Hunter-Gatherer Tech-
nology. InHunter-GathererEconomyinA-e-history, edited by G.
Bailey, pp. 1 l-22. Cam-bridge University Press, Cambridge.
1989 Re-Tooling: Towards a Behavioral Theory ofStone Tools. In
Time, Energy and StoneTools, edited by R. Torrence, pp. 57-66.
Cam-bridge University Press, Cambridge.
Tsirk, A.1974 Mechanical Basis off Percussion Flaking:
Some Comments. American Antiquity39: 122-130.
1979 Regarding Fracture Initiations. In LithicUse-WearAnaZysis,
edited by B. Hayden, pp.83-96. Academic Press, New York.
Tuffreau, A.1978 Les industries acheuleennes de Cagny-la-
Garenne (Somme), LAnthropoZogie 82:37-60.
Tuffreau, A., and J. Somme1988 Le gisement PaGoZithique moyen d
e
Biache-Saint-Vaast (Pas-de-CaZais):Stratigraphie, environnement,
htudesarch&ologiques. Memoire de la SocieteP&his torique
Francaise, Volume 2 1. Paris.
Dibble and 0. Bar-Yosef, pp. 279-292. Pre-history Press,
Madison.1Shott, M.u.
1996 An Exegesis of the Curation Concept. Jour-nal of
Anthropological Research 52:259-280. Volkman
Sollberger, J. 19831977
Speth, J1972
1974
1975
On Fluting Folsom: Notes on Recent Experi-ments. Bulletin of the
Texas Archaeologi-cal Society 48:47-52.
Boker Tachtit: Core Reconstructions. In We-history and
Paleoenvironments in theCentral Negev, Israel. Volume Hl:
TheAvdat/Aqev Area, Part 3, edited by A.E.Marks, pp. 127-190.
Southern MethodistUniversity Press, Dallas.The Mechanical Basis of
Percussion Flaking.
American Antiquity 37:34-60. Whittaker, J.Experimental
Investigations of Hard-Ham- 1994 Fltntknapping: Making and
Understand-mer Percussion Flaking. Tebiwa 2 17:7-36. ing Stone
Tools. University of Texas Press,Miscellaneous Studies in
Hard-Hammer Per- Austin.cussion Flaking: the Effects on Oblique
Im-
Figures on pages 163 - 170