This ambitious book represents the most significant systematic attempt since Gustaf Aulén's Christus Victor to defend the patristic theory that Christ's redemptive death is best interpreted as a ransom to the devil. Lombardo's strategy is simple: to sketch out the moral and scriptural desiderata for any theory of the atonement; to demonstrate that the rivals to the ransom view cannot satisfy those desiderata; and finally to show that the ransom view can. Lombardo considers there to be two main alternatives to the ransom theory: Anselm’s, according to which Christ's life and (especially) death constituted some kind of satisfaction made to God in place of the punishment due to humankind; and Abelard’s, which according to Lombardo involves seeing Christ's death as the revelation of God's love for us, a revelation which in turn inspires us to act in accordance with the demands of charity. The three-fold taxonomy of theories of the atonement is revealing of the way in which Lombardo understands the nature of the ransom theory. In short, he sees what used to be called the ‘physical’ or ‘mystical’ theories of redemption – according to which God's becoming human somehow divinises humanity – as integral to the ransom theory. As Lombardo sees it, what is central is God's provoking the evils of the crucifixion so as to ‘draw out the power of evil . . . and take it upon himself’; so that ‘after absorbing the full force of evil in his crucifixion, Christ overcomes death by his resurrection and makes it possible for us to share in his victory’ (p. 229).

It seems to me that Abelard's theory, as presented here, is not a serious theological contender. So I will focus instead on Lombardo's analysis of Anselm – not with a view to defending Anselm's theory, but merely with a view to assessing the cogency of Lombardo's criticisms. According to Lombardo, the key moral requirement of any theory of the atonement is that the theory does not involve God's intending moral evil – for example, the death of someone undeserving of death. Employing the principle of double effect, Lombardo maintains that a foreseen side-effect of an action is intended only if it is the case that the side-effect is logically implied by the action itself (i.e. such that the action cannot be coherently described apart from the side-effect). According to Lombardo, Anselm's theory involves God's intending Christ's death – i.e. the death of someone undeserving of death. The reason is that, according to Anselm, God willed the following three things in the incarnation. First, he willed Christ to persevere in justice. Second, he was ‘not willing for the world to be restored unless Christ endured his passion’. Third, he ‘positively willed for the human race to be restored’ (p. 152). The conjunction of these three claims entails that God ‘positively willed for his Son to endure death’ (p. 152). Lombardo highlights what he takes to be Anselm's unsuccessful attempts to resist this conclusion in Cur deus homo 1.8–10, where Anselm gets no further towards a response than showing that God did not coerce Christ to undergo the passion.

A lot, of course, depends on what it is for one action logically to include a set of effects. Here the crucial divine ‘intention’ is the second one: being unwilling that the world be restored unless Christ endures his passion. Anselm's actual view is that, given the contingent circumstances of fallen humanity, there is no other way for God to redeem human beings; and it is only in this restricted sense that God is unwilling that the world be restored in any other way. But this is far from claiming that the things that God actually intends – Christ's persevering in justice, and the restoration of the human race – logically include Christ's passion and death. Far from it: Christ's death is contingently unavoidable, and is thus a classic case of an unintended side-effect of God's intentions. The discussion in Cur deus homo 1.8–10, is not intended to answer Lombardo's problem (which Anselm, I think rightly, did not see to be a difficulty). Its aim is merely what it is stated to be by Anselm: to show how non-coercion is sufficient for freedom (i.e. to defend what we would now label a ‘compatibilist’ account of free will in the context of Christ's life and death).

If I am right, the book does not successfully defend the ransom theory. But this in no way detracts from details of the theory as Lombardo develops it, or constitutes an attack on the theory itself. Lombardo shows the theory to be a worthwhile and live option in the array of views surrounding one of the more intractable loci in Christian theology.