>For example, Neil argues for not demanding a strictly scientific (true/false)
>basis for our proofs, lest we discard hypotheses that may in fact be true. He
> is absolutely right. In astrophysics for example, in spite of nearly a half
> century of study by thousands of astronomers, we still don't know for certain
> whether the universe is open (expanding forever) or closed, (a gradually slow
>ing expansion which will ultimately stop, then collapse on itself). Astronomer
>s get by with probabilistic causal chains of logic, which are judged by the ag
>gregate probability they represent. On that basis, Fred Hoyle's "steady state
>universe" hypothesis has been rejected, while astronomers still study the open
> and closed universe hypotheses. Although the aggregate probabilities slightly
> favor the closed universe, the open universe hypothesis is still viable. It'
>s an technique that might work for us as well. (more later)
As your example illustrates, science does not expect every theory to be
strictly true or false.
Even mathematics is to a large extent a matter of consensus. There was
a famous case of a well-known theorem that was accepted by the entire
worldwide mathematical community for many years before someone pointed
out a small flaw in the proof.
Any science that deals with the phenomenal world can't even be that certain.
For one thing, today's accepted theory (Newtonian mechanics) may turn out
to be a special case of tomorrow's larger theory (Einstein's special theory
of relativity). In hard sciences at least there is the concept of
falsifiability, which is that theories are preferred that are at least
capable of being falsified, because that means they are capable of being
tested.
For another, many subjects are not describable with neat equations.
History is one of those subjects. Despite all the grand cyclical
theories of Spengler and Toynbee, despite all the sweeping analyses
of Kennedy and Braudel, history happened in a haphazard way and left
varying degrees of evidential debris. Like a lawyer in a courtroom,
a historian has to make a case based on whatever evidence is available.
Except in the historian's case, much of the evidence is long cold,
some of it may have been deliberately suppressed generations ago,
and almost certainly most of it is lost.
Another grand historical work of recent years, Plagues and Peoples
by William Hardy McNeill, asserts many theories, such as that less
organized peoples who came into contact with civilized groups were
almost certainly killed off repeatedly by diseases that were endemic
to civilized people but that were epidemic to the uncivilized. He
also repeatedly notes that there is little or no chance of ever
finding much written evidence of this phenomenon, because peoples
that considered themselves civilized didn't write about illnesses
of barbarians. Yet that doesn't mean it didn't happen. And he makes
a good case that it must have happened, because it fits into everything
he *does* have evidence about. It's hard to say how such a theory could
even be tested via written historical evidence, even though it is important
and quite possibly true. Fortunately, there is some chance that it might
be testable via archaeological evidence if the state of the art advances
far enough.
In the end, evidence counts, and the more the better, but what makes
the most difference is whether it fits into a coherent scheme.
This commentary is not meant to contradict Joe Erkes' comments in any way;
I'm merely approaching the same topic from a different direction.
John S. Quarterman <jsq@quarterman.org>