26 BASICPRINCIPLES – MAKING THEMOST OF THECLUES
statistical confidence of one standard deviation,
meaning that there is a 68 per cent chance of
the actual date lying within the range 1664 –
1864 years ago. Higher levels, such as two standard
deviations, meaning a 95 per cent chance, or three
standard deviations, meaning a 99 per cent chance,
mean doubling and tripling the limits respecti-
vely, so the example above would be ±200 or ± 300
years. Therefore, you can have a better chance of
being within the limits, but the limits become
wider.- Radiocarbon dates can be quoted as either before
present (bp, present being taken as 1950) or in cal-
endar years (i.e. bcor ad). - If the radiocarbon date has been calibrated (re-
assessed in relation to another dating system such
as dendrochronology) then this is often indicated
by adding ‘cal’ (i.e. ‘cal BP’). - Methods of measuring and calibrating dates also vary
between different laboratories. The name of the
laboratory and the methods used are therefore
likely to appear with published dates.
It is important that factors such as the meaning of the
±statistical confidence, the method of calibration and
the actual radiocarbon technique used are all taken into
account and fully understood when using radiocarbon
dates. Recommended procedures for radiocarbon sampling
are given in chapter 15.
Historical association: While the above techniques
are relatively complicated methods of gaining an absolute
date, the comparatively simple technique of using writ-
ten records should not be overlooked. Written records
allow historical association to provide archaeological
clues with absolute dates.
While historical dates seem very attractive they do
have their problems. Reliable written documents only go
back a relatively short time into our past. Where they
do exist, it is sometimes possible to misuse them or look
to them for easy solutions. For example, written records
provide information about the Spanish Armada but the
fact that the event occurred and is recorded does not mean
that every sixteenth-century vessel found in UK waters was
associated with it. In addition, historical documents are
the record of witnesses, some of whom may be biased or
simply ignorant. Their accuracy should not be taken
for granted. In the present day, reading a selection of
daily newspapers can reveal conflicting interpretations of
the same event, and the same can be true of other his-
torical documents. Archaeological dates should not be
ignored simply because they conflict with documentary
evidence.
In addition, as with all dates, direct historical associ-
ations relate to one point in the development of the site- in the case of a shipwreck, the arrival of the material
on the sea-bed. Subsequent events on the site must be dated
independently. A problem can also arise with coins or
cannon, which appear to give a clear date and historical
association because of the inscriptions on them. A coin
found on a site may commemorate a particular ruler, the
period of whose reign is known from historical sources.
This may well provide a good absolute date for the mint-
ing of the coin, but it gives little secure information
about the date of the coin’s loss or burial. Clearly it
cannot have been lost before it was minted (this date is
called a terminus post quem), but it could have been lost
a long time afterwards.
Sources of further information on these and other
absolute dating methods are given at the end of this
chapter.
Relative dating
Relative dating can only indicate whether one process
occurred before or after another one. It cannot reveal
the length of time between the two events and neither
can it provide a date in years that places the event in a
conventional timescale. However, it is very useful for
determining whether the information was deposited
early or late in the development of the site, and for
providing a framework into which absolute dates can
be placed.Typological dating: The value of typology as an aid to
research has been noted above, but it also has a role to
play as a form of relative dating. The form of objects
designed to perform the same function often changes
over time. If earlier and later characteristics can be
recognized it is possible to reconstruct the sequence
of development and give each object a relative position
within it (figure 4.7). There is a real danger of such
sequences being unsound because they are based on
assumptions of early and late characteristics. Therefore,
as much evidence as possible should be introduced to
support any conclusions drawn.Stratigraphic dating: It has already been shown
how contexts can be used to identify the events in the
history of a site. It is a simple process to start to recog-
nize and study the order in which they occurred. This
gives us a sequence for the events and a relative dating
technique.
The ordering of contexts is known as a site’s stratigraphic
record. The study of this record or sequence is known
as stratigraphy. The most basic principle of stratigraphy,