Databases Reference
In-Depth Information
conduct and uniformity of word usage at all times in all contexts that go beyond my
experiences of human consistency.” Second, that “the cuneiform text in question can be
read with no ambiguity in the technical terms involved. We can test this corollary by
comparing independent translations of the (transcription of the) cuneiform record.”
31
Newton concluded that the vagaries of translation and linguistics did not allow a high
reliability to be assigned. Babylonian astronomical texts had another peculiar problem,
which was that it was unclear whether statements such as “eclipse on the first day of
the third month” were
observations
of eclipses or
predictions
, thus making them wholly
useless for this purpose.
32
Even further, he denied that most of these eclipses were total—even when explicitly
described as such. One could only distinguish between a total and annular eclipses if
one is “an expert astronomer.” It seemed that Ptolemy and other ancient astronomers
actually did not know about annular eclipses, so their word could not be trusted.
33
Much of Newton's work was rebutted by F. Richard Stephenson (a British astrono-
mer with extensive expertise in Chinese texts) and Paul M. Muller (formerly of the Jet
Propulsion Laboratory) who nonetheless acknowledged the difficulties:
Since the subjective impression of an eclipse on the part of a human observer is the
data to be used in this investigation, there is no dei nable and repeatable set of
experiments which will
prove
the conclusions we are about to give in this matter. In
that sense, this investigation must necessarily be somewhat “unscientii c.” We, and
others before us, have been convinced that at least some of the admittedly imprecise
and often emotional records retained through history by eye-witnesses to large eclipses
can, nevertheless, constitute highly reliable and usable scientii c observations.
34
The pair relied on the “the subjective impressions which arise from total, and near-total,
eclipses.” The psychological impact of totality, even compared to 0.99% of totality, was
so tremendous that it could be relied on. This also had a useful side ef ect in that they
did not have to worry about the training or knowledge of the observer.
35
They could
not rely on modern instrumentation or data collection, only raw human experience.
What was needed was to think about those ancient observers as instruments: “to make
use of these data as scientii c observations requires that we calibrate and understand the
observers as human beings, and see the kind of events in the context of what the impact
would be on these observers.”
The observers, long dead, needed to be calibrated like a thermometer. Astronomers
were hardly trained in such a task, and had to draw on various other disciplines for it
