GOD nothing cannot cause anything It must be something other than itself, for nothing can cause itself Let us call that something else A Is A itself caused? If not, it is a Wrst cause, which is what we were looking for If it is caused, let its cause be B We can repeat the same argument with B Then either we go on for ever, which is impossible, or we reach an absolute Wrst cause Scotus, like Aquinas, makes a distinction between two kinds of causal series, one of which he calls ‘essentially ordered’, and the other ‘accidentally ordered’ He does not deny the possibility of an unending regress of accidentally ordered causes, such as the series of human beings, each begotten by an earlier human Such a series is only accidentally ordered A father may be the cause of his son, but he is not the cause of his son’s begetting his grandson In an essentially ordered series, A not only causes B, which is the cause of C, but actually causes B to cause C It is only in the case of essentially ordered series—e.g a gardener moving earth by moving a spade—that an inWnite regress is ruled out An accidentally ordered series is, as it were, a horizontal series of causes; an essentially ordered series is a vertical hierarchy; and Scotus tells us, ‘inWnity is impossible in the ascending order’ (DPP 4, p 22) Even after the two kinds of series have been distinguished, there seem several weaknesses in Scotus’ argument, considered as a proof of the existence of God In the Wrst place, it seems, like the proof of the Summa contra Gentiles on one interpretation, to assume that it is sensible to talk of something non-existing as having, or lacking, the power of coming into existence.12 In the second place, it is not clear why instead of a single inWnite Wrst cause the argument does not lead to a number of Wnite Wrst causes Scotus in fact admits that he has not produced a proof of God; but the reason he gives is not either of the above Unlike Aquinas, who took as his starting point the actual existence of causal sequences in the world, Scotus began simply with the mere possibility of causation He did so deliberately, because he preferred to base his proof not on contingent facts of nature, but on purely abstract possibilities If you start from mere physics, he believed, you will never get beyond the Wnite cosmos But the consequence of this is that the argument, up to this point, has proved only the possibility of a Wrst cause: we still need to prove that it actually exists Scotus in fact goes one better and oVers to prove that it must 12 See p 203 above on objective possibility 305