When are events identical, when distinct? What criteria are there
for deciding one way or the other in particular cases?

There is a familiar embarrassment in asking identity questions of
this sort that comes out clearly if we rephrase the question slightly:
when are two events identical? Or, when is one event identical with
another? It seems only one answer is possible: no two events are
identical, no event is ever identical with another. It is hopeless to try
to improve matters by asking instead, when is an event identical
with itself? For again, only one answer is possible: always.

The difficulty obviously has nothing special to do with events, it
arises in relation to all identity questions. The only move I know for
circumventing this conundrum is to substitute for questions about
identities questions about sentences about identities. Then instead
of asking when events are identical, we may ask when sentences of
the form 'a = b' are true, where we suppose 'a' and 'b' supplanted
by singular terms referring to events.

We have no sooner to restate our problem in this standard way,
however, than to realize something scandalous about events.
Events, even in the best philosophical circles, lead a double life. On
the one hand, we talk confidently of sentences that 'describe' or
'refer to' events, and of cases where two sentences refer to the same
event; we have grown used to speaking of actions (presumably a
species of event) 'under a description'. We characterize causal laws
as asserting that every event of one sort is followed by an event of
another sort, and it is said that explanation in history and science is
often of particular events, though perhaps only as those events are
described in one way rather than another. But -- and this is the other
hand -- when we turn to the sentences, formalized in standard ways

-163-