ABSTRACT
A defi n i tion can have two dimen sions: it may artic u late the meaning and use of a word, and it can provide a tool in the construc tion of an explan a tion. The linguistic and cultural turns in philo sophy, history and anthro po logy have made many schol ars more inter ested in the former, while the social sciences have focused more on the latter. Both programmes are defens ible, but neither should be conduc ted in ignor ance of the other, as Hobbes was aware. 1
Understood as a tool, a defi n i tion aims to identify a set of criteria that serve as neces sary and suffi cient condi tions in picking out a distinct phenomenon or class of phenom ena. That phenomenon (or class), once distin guished, can be under stood or explained by refer ence to sets of ante cedents or causal condi tions. If we confl ate the defi n i tions with the causal conditions we get tauto lo gies. For example, we cannot explain the fact that increas ing numbers of young men are failing to marry by saying that they are bach el ors. Nor do we explain the fact that people are using their offi ces to gener ate income for them selves by saying that they are corrupt. That may be a rede scrip tion of the state of affairs, but it is not an explan a tion. Explanations are in some thing of a state of tension with defi n i tions. Ideally, one’s class of events (picked out by a defi n i tion) is a class that also has some explan at ory unity, in the sense that members of the class are suscept ible of the same causal explan a tions. Yet we do not want
to identify the class by the fact that the same explan a tion works, since we need to pick out the criteria that unite the class inde pend ently from the condi tions that explain its occur rence.
