Archaeology Magazine - January/February 1986 - 50

"the odds against both structures
[Stonehenge I and III, different
structures of different periods in each
of which Hawkins found astronomical order] having the alignments
by accident are ten million to one."
Unfortunately, statistics used this
way mean nothing at all. A simple
example can illustrate this point. I
was born in Bradford, a mediumsized city in northern England, on
October 13, 1951, and I am one of
20 archaeologists on the faculty at
Cambridge University. Some simple
arithmetic shows the chances of
this happening to someone born on
that day are a staggering 250,000
to 1 against, if you reckon only
Britons in the sums. But some of
our faculty were born abroad.
Allowing for non-Britons takes the
odds up toward 500,000 to 1. 1 work
with a colleague who was born in
the little town of Kimball, Nebraska,
on March 3, 1956; the chances that
someone should be born exactly
there then who ends up researching
archaeology at Cambridge are
somewhere near 2,000 million to one,
depending on your assumptions.
So the combined odds against the

two of us working together are
the two figures multiplied together,
or a truly astronomical 1,000 million.
The astronomical probability only
concerns this particular configuration of these particular individuals
born in these particular places on
these particular
days. The
same goes for Stonehenge. It is not
enough for the archaeoastronomers
to prove that any particular Stonehenge configuration is astronomically
valid and statistically very improbable. They must also show that
astronomical configurations of that
general class are improbable as
a whole.
Here is the nub of the problem.
Possibilities are explored, hypotheses developed, and astronomical
probabilities calculated because
Stonehenge suggests them. But
that very fact corrupts the test of
their validity. A scientific hypothesis
developed on one set of data is
normally tested on another independent set of data. If it works there,
it is validated. A hypothesis developed on one Stonehenge should
be tested on another!
Prehistory, alas, is not like that.

Most impressive of the
astronomical alignments
at Stonehenge are the
eight made by the four
"station stones" (91-4).
Sceptics feel that the evidence for astronomical
links are not overwhelming
even here.

Stonehenge is unique: no other
monument in prehistoric Europe
brings together all its distinctive
features. The best we can do is to
assess the fit of each hypothesis
to the evidence, and to look for a
consistent pattern across sites which
share some of its features. No one
can prove Stonehenge astronomy;
the best we can hope for is a theory
which fits both the special features
of Stonehenge and the general evidence we have for this period of
prehistory. This consideration puts
many hypotheses about Stonehenge
into a kind of limbo- judged by
archaeological opinion to be unwarranted, but not finally disproved
in the way a false theory in experimental physics can be absolutely
ruled out. A New Zealand researcher,
for example, has shown how a circle
of 56 pits from the early phase at
Stonehenge known as the "Aubrey
holes" can be used to predict tides
in the English channel, 30 miles
to the south (although it helps in
working the system to have a good
knowledge of celestial mechanics).
Archaeologists reject the idea
because ( 1 ) the Aubrey holes were
deliberately filled up soon after
they were dug; (2) advanced understanding of celestial mechanics in
3000 b.c. seems unlikely; (3) nothing
in the pattern of other Stonehengelike monuments fits with the idea;
and (4) the value of tidal predictions to prehistoric Britons seems
unlikely to have been so very great.
All these are strong objections
but neither any one by itself nor all
four (and there are more) collectively can rule out the possibility with
complete certainty.
There is, fortunately, one way out
of the impasse. Stonehenge is
unique, but it shares many features
with stone circles, rows and other
stone structures of prehistoric
Britain. So a search can be made for
a systematic pattern in these contemporary monuments: evidence of
astronomy across a whole series
of sites would clinch the argument in
a way the anecdotal Stonehenge
example cannot.
Alexander Thom, retired professor of engineering at Oxford, has

50

ARCHAEOLOGY
Archaeology Magazine - January/February 1986

Table of Contents for the Digital Edition of Archaeology Magazine - January/February 1986
