Archaeology Magazine - January/February 1978 - 9

channel which leads to a fine sieve where the
fragments are collected. This technique was
developed in a research project based in
Cambridge, England, which was investigating the
early history of agriculture. The advantages of
froth flotation are speed and efficiency; it can
process up to a ton of earth a day and recover
carbonized material even if it is denser than
water and would not have been retrieved by
conventional flotation techniques.
These new, large-scale flotation devices are as
much an advance in archaeobotany as the
telescope was in astronomy. Not only can far
more information now be obtained, but new
areas of research have been opened up. The
most important result is that one can now obtain
a far more accurate indication of the importance
of each plant food in an ancient community. At
first this task does not sound especially difficult,
since it seems reasonable that the most important
food plant is the one preserved most often or
in the largest quantities. Unfortunately, the
situation is not quite so straightforward.
Although all plant remains from archaeological
sites are preserved accidentally, their
preservation is not random. It is a simple
fact that some plants or parts of plants stand
a better chance of surviving than others.
Since roots, shoots and leaves are very rarely
preserved, for example, far less is known about
ancient vegetables than about seed or grain
resources like legumes and cereals. Some of the
perishable plants, however, may be more
important than those that are commonly
preserved. For example, cereals such as einkorn
or spelt need parching, often in an oven to free
the grains from their kernels and are therefore
more likely to be carbonized than free-threshing
cereals such as breadwheat. Because legumes
such as lentils or peas can be eaten without
parching, they are unlikely to have been
preserved in large amounts except when part of
a settlement burned down. The survival of fruits
in the archaeological record is strongly affected
by the number and size of seeds in each piece of
fruit. Clearly, fruits like raspberries or figs were
not necessarily more important than plums or
olives simply because they left several hundred
times more seeds.
Human use and preparation of plant foods
also affected their preservation. Fruits, for
example, would leave little trace if they were
crushed into juice instead of being dried and
stored; similarly, barley would be poorly

preserved if it was used for brewing. Although
the preservation of a plant food is largely
dependent upon its type and use, the actual
composition of a particular sample probably
results from human activities performed even
before the plants were stored and eaten. This
factor is especially important in the case of cereal
crops. Before grain can be eaten, several
processing activities such as threshing,
winnowing and cleaning are necessary to
separate the grain from the chaff, straw and
weeds. Some of these activities produce a
characteristic residue. Winnowing, for example,
removes the chaff, smaller grains and weed
seeds; grain cleaning removes the remaining
weed seeds and possibly a few of the actual
grains. It is obviously important not to mistake
the crop for the part that was deliberately
removed from it, nor to fail to recognize the
processing stage at which a particular sample of
a cereal crop was accidentally preserved. Both
errors can lead to inaccurate assessments of the
plans importance to the community.
T he first task in investigating a setdemens use
of food plants is to identify the plant processing
activities performed there; plants that were
prepared for consumption can then be
distinguished from those that were discarded. In
my own research, I have tried to make these
judgments by considering three variables - the
composition of each sample, its context and,
when possible, the size of the cereal grains.
Because crop processing activities such as
winnowing or cleaning alter the composition of a
sample, the composition and grain size of a
deposit can provide important clues about the
stage at which the sample was accidentally
preserved. The context of a sample can also be
informative; plant material in a storage jar, for
example, should represent plant food at a more
advanced stage of preparation than that in a
refuse dump or threshing area.
I am not, of course, claiming that every
sample of plant remains from an archaeological
site will provide unequivocal evidence of
individual processing activities. In many cases,
the samples may be too small or may represent
the results of several different activities. If,
however, several samples collected from a large
area fall into a small number of categories, with
distinct regularities in composition, context or
grain size, then it is reasonable to assume each
9
Archaeology Magazine - January/February 1978

Table of Contents for the Digital Edition of Archaeology Magazine - January/February 1978
