An explanation is some type of answer to the question `Why did happen?'.
It can contain any or all of the following characteristics:
``What precise sequence of events and interactions caused
to
happen in this situation?'' We argue that we can reduce the problem to its
constituent interactions, all of which we can explain (often in a precise,
quantitative fashion). Thus we have explained the connection between our
initial state and the event
in a causal fashion. This sort of
explanation can be used either as pre- or post-explanation of the
event
.
``What were the fundamental details of the initial state which
caused
?'' This allows us to generalise and determine some
characteristics of the class of initial states which bring about
, as
compared with the space as a whole. If we can't generalise, we may reach
the conclusion that we were dealing with a special case.
``Assuming this situation isn't special in any way, why did
,
rather than
happen?'' This is an elaboration of the previous
point. Presumably we can imagine a number of different outcomes to which
the system could progress (If we're so sure it'll reach
, why aren't we
researching something more interesting). Therefore we would like to have
some form of explanation for the mapping between different classes of
initial state, and the different outcomes