In this chapter we address an issue that is a difficulty for the Conserved Quantity theory; and not only for the Conserved Quantity theory, but also for many approaches to causation. We offer a solution that is available not only for the Conserved Quantity theory, but for most theories of causation.

We might be tempted to think that preventions, such as 'the father's grabbing the child prevented the accident,' and cases of causation by omission, such as 'the father's inattention was the cause of the child's accident,' are examples of causation. Such cases are 'causation' by prevention or omission, and they almost always involve negative events or facts as one or both of the relata. I will call this relation 'causation.'

For example, in 'the father's inattention was the cause of the accident,' the effect is a real occurrence, but the cause is an omission, a failure to do something. On the other hand, in ' his grabbing the child prevented the accident,' the cause is a real occurrence but the effect is the nonoccurrence of something. The former is omission, the later prevention, but in both cases we have causation, and both cases involves negative events or facts. In both cases, we can recognise that it is not literal causation, yet we still want to use the term 'cause.'

In general, suppose we have a case of singular causation where A causes B. If A is a negative event or fact, then we have a case of causation by omission; if B is a negative event or fact, then we have a case of causation* by prevention.