The variables can be divided into five groups; raw material characteristics,
edge morphology, macroscopic edge wear, microscopic edge wear and polish
distribution and characteristics. These variables and the recording sheet
developed for use with this method are illustrated in Figure
37.
5.1 RAW MATERIAL
The raw material is described in respect to its grain size, surface topography
and topographic features. These are characteristics which have been shown
to affect the appearance of wear traces on stone tools under the microscope,
but are not diagnostic of function.
5.1.1 GRAIN SIZE
Grain size is recorded according to the visual appearance of the tool, and
often relates to the colour of the stone. For example, the black areas of
Brandon flint are usually fine grained, the grey areas tend to be medium
grained and the white area coarse grained. This estimation of grain size
is made on the basis of comparing flints from different sources and also
the grey tone of the flint is often a function of the way in which light
is reflected by the differing grain sizes. It is important to record grain
size because it can significantly effect the visual appearance of the flint
under the microscope. Sometimes the effect is such that coarse grained flint
makes the observation of polishes almost impossible, because the level of
reflectiveness produces a blurred appearance.
5.1.2 .TOPOGRAPHY
An undulating or ridged topography of the flint surface away from the polished
area can affect the distribution of the polish. Therefore the topography
is recorded as it may correlate with polish distribution variables, and
so indicate that the distribution is not functionally determined, but a
product of the original topography of the unused surface. (see
SECTION 6.5)
5.1.3. TOPOGRAPHIC FEATURES
These features often explain the origin of the topography. Percussion ripples
are produced during the conchoidal fracturing of the blank and radiate concentrically
from the point of percussion. Such percussion ripples often produce an undulating
topography (see Figure 38).
Edge feathering often occurs on thin edges and is a product of the process
of detachment of the blank from the core resulting in overlapping fracture
planes that have a feathered appearance. This often results in a ridged
topography on the edges of the tools (see
Figure 39).
5.2 EDGE MORPHOLOGY
Aspects of the edge morphology are measured metrically and refer to the
working edge of the tool. The variables are edge angle, edge length, edge
profile, thickness and the shape of the tool. These variables were chosen
from a larger group of possible variables, because in a previous study these
particular variables were the ones which seemed to provide reliable functional
information (Grace 1981). Potentially
used edges are identified according to the following definition. Used edges are created by retouch or utilisation or
may be natural edges opposed by backing retouch, as with a backed knife.
The ends of the edges are determined by the limits of the retouch or utilisation,
or where a continuously retouched edge turns through a sufficiently large
angle to indicate the cessation of the worked edge. The criterion of a change
of direction of the edge line of more than 40 degrees within a distance
of 1 millimeter is used, as suggested by White
(1969, 24).
Retouched or utilised edges on the same piece are considered in two ways:
one, as lateral (to the used edge) retouch for shaping the handle, or manufacturing
traces (eg. a truncation or a stop notch on a burin) and two, as distinct
and separate working edges. This procedure is necessarily somewhat subjective,
but the nature and placement of backing or piece-modifying retouch is usually
sufficiently distinctive from retouch designed to produce a working edge.
Unretouched pieces that have very little edge damage so that the evidence
of utilisation is marginal, can be examined by looking at the relationships
between the morphological attributes of such edges. The morphological variables
can be used to estimate the functional capability of edges, and can be used
not only for functional interpretations (see
section 6), but also to operate as feedback to the initial problem of
determining whether a piece could have been used or not. For example, the
bulbar end of a flake could be described as a short, obtuse angled, thick,
convex edge of a short flake, and would not normally be an edge suitable
for use.
Pointed tools require some explanation as they represent an exception to
the above criteria. A subjective assessment has to be made as to whether
the converging retouched edges represent two distinct working edges, or
the convergence was deliberately produced for functional reasons; i.e.,
the convergence was designed to form a point. This problem corresponds with
the typing of a tool as a Mousterian point or as a convergent scraper by
the thickness of the cross section of the tip (Bordes
1961). Bifacially retouched points, where the retouch is invasive are
more obviously designed as points. The overall morphology of a projectile
point, for example, is indicative of the tool's potential function. The
morphological attributes of points are measured slightly differently to
other tools. The working edge is taken as both of the edges that converge
to a point, so that the middle of the working edge is the apex of the point.
Worked edges thus isolated are marked on the drawing and taken as the base
from which to record the morphological attributes. This procedure allows
for the selection of potentially worked edges prior to the microscopic examination.
Microscopic examination is then used to confirm which edges were actually
used, by the presence of micro-edge wear and polish. Though all edges and
surfaces of the tools are examined microscopically, the above procedure
has proved reliable in isolating worked edges.
5.2.1 .EDGE ANGLE
Edge angle is defined as the angle between the ventral surface and the retouched
or utilised edge (assuming direct retouch), and corresponds to Wilmsen's
edge angle (Wilmsen 1968, 985),
and Movius and Brooks angle BAC (Movius
and Brooks 1971, 269). Edge angle is not the same as the spline plane
angle as described by Tringham et al.
(1974). In the case of inverse or bifacial retouch, the edge angle is that
of the working edge created by such retouch. Though there can be a variation
of the angle along the length of an edge, this is generally slight and measuring
the angle at the mid-point of the working edge has been found to be consistently
representative (Figure 40).
As described above the middle of the working edge of points is defined as
the apex of the point, so that the edge angle of points is measured as the
angle created by the convergence of the two edges. The edge angle measurements
are taken with a goniometer (Figure
41). For a comparative study of edge angle measurement techniques see
Dibble and Bernard 1980.
5.2.2 EDGE LENGTH
Edge length is defined as the maximum length of the working edge whether
created by retouch, utilisation or a natural edge indicated by backing retouch
of the opposing or lateral edge, as explained above. The curvature of the
edge is allowed for by fitting a line (using a piece of non stretchable
plastic) or a designers curve to the edge. This can then be straightened
and measured against a ruler (Figure
42).
The edge length of points is taken from the shoulder round the point to
the opposite shoulder (Figure 43).
5.2.3. THICKNESS
Thickness is defined as the maximum thickness of the support piece taken
perpendicular to the mid-point of the working edge. Measurements are taken
with calipers (Figure 44).
The thickness of points is taken as the maximum thickness of the support
piece with the calipers placed at the apex of the point i.e. the mid-point
of the working edge as defined here for points (Figure
45).
5.2.4. EDGE PROFILE
Edge profile is defined as the shape of the working edge in plan, which
may be convex, concave or straight as indicated by the ratio of the perpendicular
measurement divided by the chord. The chord is the linear distance between
the extreme ends of the working edge. It is measured by aligning the ends
of the working edge along the y axis of millimeter graph paper and reading
off the measurement from the graph (Figure
46).
The perpendicular is measured by taking the maximum distance between the
working edge and the y axis. The measurement is read off from the x axis
Concave profiles are given a negative score and by dividing the perpendicular
by the chord, straight edges have a value of 0. The .edge profile of points
is calculated with the base of the point placed on the y axis (Figure
46). For examples of profiles see Figure
47.
5.2.5. SHAPE
The shape ratio is calculated by dividing the length by the height. Length
is defined as the maximum lateral dimension of the piece with the working
edge as the base. It is measured by orienting the working edge along the
y axis of the graph paper and reading off the value (Figure
48). Height is defined as the maximum vertical dimension of the support
piece with the working edge as the base. Measurement is by reading off the
value from the x axis when the piece is oriented as for length The calculation
of the shape ratio means that a long thin support piece, such as a blade
supporting an end scraper, will have a relatively low value, compared with
a flake supporting a side scraper. A blade with a lateral retouched edge
will also have a high value, so that this measurement will not, and is not
intended to, distinguish between blades and flakes, as it depends on the
position of the working edge in relation to the support piece. It measures
the shape of the part of prehension while in its working attitude. The shape
of points is taken with the apex of the point placed on the y axis (Figure
48). For examples of shapes see Figure
49.
The remaining variables are recorded separately for the dorsal and ventral
surfaces. The presence of intentional retouch used to form the tool is most
common on the dorsal surface, being initiated from the flatter ventral surface.
Retouch can often make the recording of edge wear and polish variables difficult,
and sometimes impossible in the case of edge wear. Simply, it is often not
possible to differentiate between small flakes detached during intentional
retouch and those detached by use and often the dorsal surface is recorded
as retouched without further comment.
It is essential to record polish variables both for the ventral and dorsal
surfaces as the presence of bifacial polish obviously demonstrates that
both surfaces were in contact with the worked material. Also the presence
of edge wear on one side and absence on the other is a good guide to whether
the motion is uni- or bi-directional.
5.3 EDGE WEAR
The third and fourth groups, macroscopic and microscopic edge wear, describe
any modifications to the edge other than that from intentional retouch.
These two categories include the presence, amount and type of fractures,
the presence and degree of rounding of the edge, and in the macroscopic
category, the presence or absence of visible gloss on the working edge.
MACRO-EDGE WEAR
5.3.1 AMOUNT OF FRACTURES
Macro fractures are those seen by eye. The number of fractures is recorded
as greater than, or less than 5 fractures per 10mm. of working edge. This
is a simple way of quantifying minor edge wear which may not be significant,
and a greater amount of edge wear that may be functionally diagnostic. A
value of less than 5 fractures per 10mm. may result from mild accidental
damage or be produced by retouch, as even when retouch is unifacial some
small flakes may be detached on the opposite surface.
5.3.2 FRACTURE TYPE
Flakes fractures are produced by the normal conchoidal fracture of flint
initiated by pressure or percussion against one surface of the edge of a
tool, and leave a flake scar that is the negative of the detached flake
(Figure 50). Snap fractures
appearing as crescentic or half moon shaped fractures, occur when the edge
of the tool breaks off under bending stress and leaves no negative scar
(Figure 50). Step scars terminating
abruptly in hinge fractures leave a typical scalar negative scar and are
often produced by percussion more directly on the edge, rather than against
one side as with flake fractures (Figure
50).
Before recording edge wear, fractures have to be interpreted so as to classify
the fractures as edge wear; produced by use, or edge damage produced by
natural processes. Previous attempts to classify fractures have been based
primarily on the patterning of the fractures. It has been claimed that there
have been "studies that demonstrate a clear difference between damage
produced by utilisation and that caused by other factors" (Odell
and Odell-Vereeken 1981,90). The main study referred to is that by Tringham
et al. (1974), which differentiates wear from damage by saying that
damage from various sources produces a random pattern of fractures as opposed
to the more regular patterning caused by use. These results are based on
mechanistic mechanical experiments of considerable duration, simulated experiments
where tools have been used to carry out a task, rather than being used for
a set number of strokes, do not always produce regular patterns of edge
wear fractures and sometimes produce no fractures at all. For example test
tool number 50 (see SECTION 8) had
only macro snap fractures but these were not patterned regularly, being
separated from each other at random intervals with few areas of consecutive
fractures. Use of a tool on soft materials such as meat often produces no
edge wear, or only a few randomly spaced fractures. Even use on a hard material
such as bone can produce little or no edge wear if the working edge of the
tool is robust like the facet of a burin. When edge wear is patterned it
may be differentiated from edge damage with fractures produced by such things
as soil movement, spontaneous retouch, trampling or box damage caused by
inadequate post-excavation storage.
Edge wear that does not have consistent patterning with consecutive fractures
can be easily confused with edge damage, if the fractures only are considered.
The fractures have to be associated with an edge having the functional capability
consistent with such fractures. Edge damage usually occurs on any edge of
the tool capable of sustaining fractures. The guide is whether the fractures
are only present on a potential working edge and absent elsewhere on the
tool, an unlikely situation if the fractures were the result of damage.
However, such coincidences may occur and corroborative evidence that the
tool was utilised is required in the form of the presence of rounding and/or
polish.
Recognising edge wear on retouched tools is much more difficult. It is claimed
that edge wear can be seen on a retouched edge because it "tends to
nick, crush or abrade those part of the larger scars [that is retouch scars]
that occur between impact or pressure points (Odell
and Odell-Vereeken 1981,96). This kind of micro-fracturing however,
can be concomitant with the larger retouch fractures as a result of the
same blow; a kind of micro spontaneous retouch. Keeley quotes a case where
a Clactonian notch had small flakes scars at the centre of the notch, but
these flakes were detached by the same blow that made the notch (Keeley
1980, 27).
Odell, though claiming that retouch can be differentiated from edge wear,
admits that it is difficult, as in his blind test he states "All of
the incorrectly assessed implements were judged to have been utilised on
a substance harder than they actually were. Apparently some of the retouch
was [authors italics] mistaken for use wear and the resistance of
the worked material assessed was thereby exaggerated" (Odell
and Odell-Vereeken 1981,118). As mentioned in SECTION
4 , in this method a retouched edge may not be observed and the retouched
surface is not assessed. Only fractures on the unretouched (usually ventral)
surface are considered.
Other workers have mentioned the various problems in separating wear from
damage. Kamminga comments "It is such a delicate and subjective distinction
at times that some researchers decline to clearly divide them. There is
a good deal of sense in this caution" (Kamminga
1982). Both these authors (Odell and Kamminga) use low power techniques
which rely principally on edge wear. Workers using high power techniques
tend to diminish the role of edge wear as they concentrate on polish identification.
It is extremely difficult to determine whether or not edge fractures are
due to use or damage, especially on retouched edges. The following criteria
should be used as a guide.
1) Patterning of fractures: When a definite pattern of consecutive fractures
occurs on a potentially used edge, this indicates wear. However an unpatterned,
random oriented fracture pattern does not necessarily mean that it must
have been produced by damage.
2) Placement of fractures; If fractures occur on one edge only, and on an
edge that is potentially a used edge (determined by edge morphology) but
does not occur on other edges of the tool, then this is a strong indication
that the fractures are due to use. Edge damage would tend to occur on any
edge.
3) Corroborative evidence: To assign fractures as definite edge wear, corroborative
evidence is required in the form of polish, striations, linear features
or rounding.
4) Retouched edges: Retouched edges must be treated with extreme caution
and often only fractures on the unretouched surface should be considered.
5.3.3 ROUNDING
The presence of rounding of the edge is recorded as light or heavy which
is naturally somewhat subjective, but the two values can be related to experimentally
produced examples. Heavy macroscopic rounding can usually be felt by the
fingers as a significant smoothing and blunting of the normally sharp edge
of the tool.
5.3.4 GLOSS
Gloss refers to the presence of polish that is visible to the naked eye,
and is recorded as present or absent.
5.4 MICRO-EDGE WEAR
The same variables and values are used for micro-edge wear which are recorded
by observation through the microscope at 200 magnifications. The number
of fractures is recorded as less than, or greater than, 5 per 5mm as being
a more appropriate level of significance at this magnification. The fractures
are recorded as fractures visible only through the microscope, excluding
large fractures that would have already been recorded as macro-edge wear.
Therefore it is possible to have macro-edge wear with micro-edge wear absent.
Conversely it is possible to have micro-edge wear without there being any
fractures large enough to be recorded macroscopically. One quick guide to
recording micro-rounding is that when the edge of the tool is horizontal
under the microscope a rounded edge can be detected by the difficulty of
focusing on the very edge. This is because the rounding produces an edge
that cannot be brought within the depth of focus of the microscope at 200
magnifications. Also, if the tool is observed edge on, a rounded edge will
appear as a flat area rather than the thin line of an unrounded edge. For an example of heavy micro-rounding see
Plate 9.
5.5 MICRO-TOPOGRAPHY OF POLISHED AREA
The micro-topography of the polished area is recorded not as a variable
diagnostic of function, but because it may affect the distribution of polish
on the natural surface of the flint. For example, if the area is ridged
due to the way the flake pulled away from the core, and at 200 magnifications
the surface of the flint is seen as a series of ridges and troughs, then
the polish tends to be distributed along the ridges. This produces the appearance
of a linear polish distribution which may be mistaken as diagnostic of the
motion of the tool, whereas it is only a product of the original surface
structure of the flint before use (see
Section 6.5).
5.6 POLISH DISTRIBUTION
Polish is defined as a visible alteration of the natural surface that increases
its reflectivity. This excludes residues which are additive to, rather than
an alteration of the surface. Polish distribution is recorded as continuous
or intermittent. Again, the original topography of the tool may be responsible
for the distribution and so the polish distribution is recorded merely as
continuous or intermittent so that this can be correlated with the topographical
features already recorded. (Also see
Section 6.5).
5.7 POLISH DISTRIBUTION TYPES
Polish distribution types are schematically represented in Figure
51. These distribution types account for most of the variation produced
by our experiments using tools in simulated prehistoric tasks on a wide
range of materials. The types are simply descriptive of where, and in what
arrangement, the polish is distributed on the working edge.
A (away from the edge). Where the polish is predominantly
distributed in a band not on the very edge of the tool, but away from it
(Figure 51a, also see
Plate 7).
B (gapped). Where the polish is distributed in a band away from the edge
and on the edge, with an unpolished area in between (Figure
51b).
C (edge only/even). Where the polish is distributed evenly along the working
edge (Figure 51c).
D (edge only/asymmetric). Where the polish distribution is along the edge
but is asymmetric, in that the polish is more invasive along some portion
of the edge (Figure 51d).
E (differential). This distribution is when two different levels of polish
development are present on the same edge (Figure
51e).
A working edge may have a combination of polish distribution types, for
example see Tool 34 (SECTION 6.6
and Figure 60).
5.8 INVASIVENESS
Invasiveness is a measure of how far the polish extends away from the edge.
The measurement is recorded as less than or greater than half a diameter
of the field of view through the microscope at 200 magnifications (Figure
52). This represents distances of < 100 µm. (edge only), between
100 µm. and 400 µm. (<0.5D), and more than 400 µm. (>0.5D).
5.9 LINEAR FEATURES
Linear features are defined as lines of polish, and their orientation to
the working edge is recorded as parallel, perpendicular or angled (Figure
53 ).
5.10 STRIATIONS
Striations are scratches or grooves in the polish (Figure
54), and are recorded in the same way as linear features.
5.11 POLISH DEVELOPMENT
The last category, polish development, is broken down into the schematic
representations seen in Figure 55.
The distribution types represent the most well developed area of polish
that is present on the tool. In some cases very small areas of well developed
polish are present on an edge, usually associated with a topographical features
such as a ridge. These extremes, though representing the most well developed
area, are often distinct from the more typical level of polish development,
and should be ignored when recording the level of polish development.
A: (individual elements). This distribution is defined as having polished
elements that are clearly separated from each other within a matrix of an
unpolished surface.
A+: This distribution is when the individual elements are larger, but not
yet linked together.
B: (linked). This distribution is where the polish elements are joined together,
but the majority of the observed area is unpolished.
B+: This is a linked distribution where the linkage has developed sufficiently
so that the polished area is approximately equal to the unpolished area.
C: (all over). This distribution is when the linkage has advanced to the
stage when almost all the observed area is polished.
D: (linear). A linear distribution is represented in Figure
55, and is a special case where the polish is distributed in linear
areas.