In his book The Cosmological Argument William L. Rowe considers a notable first cause argument developed by the thirteenth-century philosopher and theologican Duns Scotus. As Rowe notices, Scotus' argument is entirely a priori. It does not appeal to empirically observable features of the world. The first cause argument of Duns Scotus is thus quite different from the famous a posteriori first cause arguments presented by Thomas Aquinas as part of his 'Five Ways'. Duns Scotus' argument for the existence of a first cause consists of the following three modal premises:
1. It is possible that there exists a first cause,
2. It is not possible that a first cause be produced by something else,
3. If it is possible for x to exist, then if x does not actually exist it is possible for x to be produced by something else.
The conclusion that there is a first cause can be derived by reductio. Suppose that a first cause does not actually exist. According to (1) it is possible that there exists a first cause. From (3) it follows that it is possible for a first cause to be produced by something else. This contradicts premise (2). The assumption that there is no first cause must therefore be rejected. It follows that there actually exists a first cause.
The argument of Duns Scotus seems unproblematic. The premises appear intuitively sufficiently justified. Each of them looks plausible enough to be accepted as a premise for metaphysical reasoning. So, does Scotus' argument indeed warrant us to think that there actually is a first cause?