The following is an excerpt from an article at the Infidels web site. Too often the basic methods of historical reconstruction are misunderstood or misrepresented. The sketchy nature of ancient history studies requires standards and rules of evidence to prevent our incomplete understanding of what happened becoming an excuse to disbelieve or believe what ever is desired. This is an example of how to compile incomplete facts into a workable and responsible recostruction of the past. I will post another example later.
The Argument to the Best Explanation
The Argument to the Best Explanation (ABE) is a formalization of the
most common form of historical argument, described and defended by
Apart from the obvious fact that a theory must be testable even to be
considered, McCullagh enumerates six criteria that must be fulfilled to
construct a strong ABE. In short, when we compare the "advocated theory"
with "any other incompatible hypothesis about the same subject" it: (1)
"must be of greater explanatory scope," that is, it must explain more
existing evidence; (2) "must be of greater explanatory power," that is,
it must make the existing evidence more probable; (3) "must be more
plausible," based, that is, on established general truths about the
time, the place, the context, etc., and the universe generally; (4)
"must be less ad hoc," that is, it must contain fewer "new suppositions"
that have no other evidential support apart from the fact that they make
the theory fit the evidence; (5) "must be disconfirmed by fewer accepted
beliefs," that is, it must be less challenged by existing evidence and
general accepted truths; and, finally: (6) "must exceed [on the previous
five criteria] other incompatible hypotheses about the same subject by
so much...that there is little chance of an incompatible hypothesis,
after further investigation, soon exceeding it in these respects" (p.
The reality is that for much of history, especially ancient history, it
is not common for any theory to be so successful as this. Historians
always deal in probability, but they get knee deep in uncertainties far
more than any other scientist or investigator. Thus, what is reasonable
to believe is, in general, what is ?most probable', not just what is
?practically certain', since such confidence can rarely be had for
claims about ancient history. But the ABE still serves the historian
here, too: while no theory in many cases can win on all six criteria,
very often one theory can win on enough of them, and by a large enough
margin, as to be the most credible, possibly even the only credible
theory. The relative credibility of two theories, in other words, rests
in proportion to their relative success on the criteria of the ABE. For
instance, based on an ABE one might correctly say that theory x is very
likely and theory y rather unlikely, therefore it is more reasonable to
believe x over y.