On the individuation of events

As Donald Davidson has pointed out,1 there are many good reasons for taking events seriously as concrete individuals, i.e., as numerically unique entities which have location in space and time. In the first place, both action theory and explanation seem to call for events qua unrepeatable, spatially located particulars. In excusing an action (a species of event) we frequently describe the action in a number of different ways. My daughter's eating all the brownies in the refrigerator this afternoon is the very same action as my daughter's eating the dessert for tonight's dinner. But she is blameless in that she didn't know that the brownies were the dessert. Similarly in explaining the expansion of a piece of metal, scientists may redescribe it in a number of different ways: in terms of the kind of material it is, in terms of the heat capacity of metal, in terms of molecular bonding and the motion of molecules. All this talk of redescription makes little sense if there are no individuals to be described in the first place. Moreover, as Davidson has shown, an ontology of unrepeatable events has great utility for the purposes of accommodating adverbs in predicate logic, allowing us to validate the intuitive inference from, say, 'Elspeth ate the brownies quickly under the deck at noon' to 'Elspeth ate the brown ies'. In brief, action theory, causation, explanation, and logical theory all seem to call for events which are concrete individuals, as opposed to abstract (timeless) and general (property-like).2 Nevertheless the claim that events are concrete individuals faces a major hurdle. If events are concrete individuals, then it should be possible to provide identity conditions for them;3 for as Quine has cautioned us, "No entity without identity".4 Unfortunately, however, the most widely discussed proposals for such conditions, by Donald Davidson, John Lemmon and Jaegwon Kim, are beset with serious problems. In order to better understand the difficulties involved in formulating identity conditions for events, let us briefly review these three proposals, paying particular attention to the problems commonly associated with them.