Verification of Statements on Causal Relationships in Experimental Research

Abstract

The causal relation is sometimes defined with reference to the conditions under which experiments are made. It has been pointed out on many occasions, in tentative formulations of a definition of cause, that an exception-free sequence of events is not sufficient for a causal relation to take place, nor is a statistical relationship sufficient for that purpose. Stronger relationships are necessary. One of the endeavours to formulate in what those stronger relationships consist is as follows: an event x is treated as a cause of an event y if in a given situation (definite variables being fixed) y occurs after a purposive producing of x (and if — in that situation — y does not occur if x has not been produced). In such a case x is a necessary component of a sufficient condition of y, and moreover x has been produced on purpose. In accordance with that idea, a connection between phenomena which occur under the conditions of experimental manipulation suffices for a causal nexus to be stated.