INTERPRETING THE FUNCTION OF STONE TOOLS
APPENDIX 1:TESTING FOR EFFICIENCY
Tools A1-5 were used for whittling a 3 cm. diameter hawthorn branch into
a point. No attempt was made to control stroke length, contact angle or
pressure exerted on the tool. Rather the tools were used in the most efficient
way possible to achieve the desired result. All the tools were used for
10 minutes except for tool A3 with which the task was completed in only
seven minutes. The tools are presented in order of efficiency as assessed
by the resulting points on the wooden branch. Morphological variables of
the edges of the tools; angle, length, thickness, profile and shape were
measured (see Chapter 4, section 5
for a definition of the variables and a description of how they are measured).
From this kind of information the effect of variables on efficiency can
be assessed. Some tendencies can be detected. An increase in efficiency
is achieved with the more acute angled edges, and with the shorter edges.
Thickness does not correlate with efficiency but as thickness refers to
the support piece it would not be expected to be a sensitive diagnostic
variable for tools used for whittling. The profile of the edges does correlate
with efficiency, the convex edges being less efficient that the concave
edges, particularly with the extreme concavity of the Clactonian notch (tool
A3). From using the tools the increase in efficiency of concave tools for
this specific task was due to the concavity fitting the piece of wood and
preventing the force being applied from being dissipated by the edge sliding
over the surface of the wood rather than cutting into it.
The shape variable does not appear to correlate with efficiency, though
the longest/narrowest support piece (tool A5, shape = 2.88), may have been
more efficient except for the denticulation of the edge which decreased
the tools efficiency as it increased the edge angle (70o). The serrated
edge was not suitable for a slicing action and tended to gouge the wood
rather than cutting it cleanly. The next longest/narrowest tool (shape =
2.14) was A3 the Clactonian notch, whose handling properties (i.e. the relationship
between support piece and the used edge) were excellent allowing a firm
grip of the tool evenly around the working edge.
Though tool A3, by far the most efficient tool, had a relatively high edge
angle (52o) the concavity of the notch (profile = -0.3) and the fact that
the unretouched edge provided a clean cutting edge more than compensated
for any deficiency due to the middle range edge angle. In fact all the unretouched
edges (A4,A2,A3) proved more efficient than the retouched tools (A1,A5).
Tools A6-9 were used for sawing the same hawthorn branch for 10 minutes.
The more acute angled edges tended to be more efficient. The relative inefficiency
of tool A7, having an edge angle of 29o, was caused by the thickness of
the support piece preventing the thin edge from penetrating into the wood.
The most efficient tool (A9) had an acute angled edge (22o) and was also
thin (thickness = 7). No concave profiled edges were used as this precludes
their use as sawing tools as it is mechanically impossible to saw with a
Clactonian notch, for example. There is a strong correlation between support
piece shape and efficiency (i.e. the longer/narrower tools were more efficient).
Tool A6 the transverse scraper, had invasive stepped retouch that not only
increased the edge angle (82o) but also created a roughened surface that
produced a wide, shallow cut. The unretouched edge of tool A7 became blunt
quite quickly because the bending stress caused by the sawing motion created
large snap fractures presenting a blunted edge to the wood. The denticulation
of tool A8 proved more efficient but the increase in edge angle produced
by the large scars detracted from its over all efficiency. Tool A9 though
typologically a side scraper proved most efficient, having an acute angled
edge (22o) and a thin, (thickness = 7) long and narrow support piece (shape
= 2.18). The fine retouch on this tool did not significantly increase the
edge angle but provided a more stable edge than that of the unretouched
edge of tool A7.
This small scale .i.simulation experiments; experimental program indicates
that there are correlations between the efficiency of tools, used for a
specific task, with their edge attributes. The techniques presented here
have been applied to a sample (n=252) of archaeological flints and the correlations
between variables used as the basis for clustering the tools. The inference
being that tools that cluster according to these variables are suitable
(i.e. more efficient) for specific tasks (Grace
1981).
Amount of work achieved after 10 minutes sawing
In contrast to the experiments discussed in chapter 1, where the function
of tools is assumed, these kind of experiments are designed to define the
limits of a tool's capability. This in turn can indicate possible functions,
or at least limit the range of functions that the tool is capable of.
CONTENTS