Thomas Holden's goal in this book is to demonstrate that Hume took himself as having provided ‘a decisive case against the being of a god with moral attributes’ (2). In Holden's view, Hume should not be understood as claiming merely that we cannot know whether God has moral attributes, or that it is logically possible, but nevertheless improbable, that God has moral attributes. Rather, claims Holden, Hume is knowingly committed to the much stronger claim that ‘no such being could possibly have a moral character’ (3). Whether or not there exists a first cause or organizing principle of the universe, such a being stands entirely outside the moral order and ‘is not a proper object of moral assessment or evaluation one way or the other’ (8). Hume is thus what Holden terms ‘a strong moral atheist’.

Given the existence of passages where Hume explicitly rejects cosmological and theological speculation as futile, it might seem that Holden's thesis should be rejected out of hand. Holden argues, however, that such passages ‘are best interpreted as moments of rhetorical grandstanding’ (45). In Holden's view, although Hume rejects any project of core natural theology, namely the drawing of conclusions regarding the deity's intrinsic nature (45), he does allow for liminal natural theology, namely ‘an aprioristic natural theology of the via negativa stripe, maintaining that the first cause simply cannot have certain combinations of properties’ (33).

Holden argues that Hume's liminal natural theology and its devastating implications for traditional theistic belief become apparent once his philosophy of religion is examined in conjunction with his views concerning human nature, particularly his sentimentalist theory of moral judgment. Hume grounds morality in sentiments of approval and disapproval. As humans, we typically react to certain character traits with approval, naming them virtues, and to other character traits with disapproval, naming them vices. Holden claims that this implies that morality is thus restricted to objects capable of eliciting a human sentimental response; objects which do not provoke such a response must be taken as having no moral nature. He finds textual support for the claim that Hume views the first cause or designer of the universe as ‘beyond the natural range of all of our passions’ (59). Thus,

Leaving aside the issue of whether we take ourselves in making moral judgments simply to be describing our emotional state, it might be objected that religious believers do take God as a natural object of human sentimental response and thus it cannot be insisted, even on Hume's view, that the deity has no moral nature. Holden's response to this objection is to note that, although Hume sometimes diplomatically seems to allow such a possibility, he more typically pathologizes such claims, regarding those who make them as ‘unbridled and self-deluding enthusiasts’ (63). Hume thus holds an ‘error theory’ of religious affection; ‘His basic claim is that in most (if not all) cases of religious sentiment, the believer's passions are not really directed toward the transcendental first cause of all, and if the believer thinks otherwise, then he must be misinterpreting the objects of his own feelings’ (74).

Holden seems to view strong moral atheism as a logical consequence of Hume's sentimentalist theory of moral judgment. He also discerns in Hume a probabilistic argument for strong moral atheism. He argues that on Hume's account of moral motivation,

To the charge that such a speculative inference is at odds with Hume's oft-stated agnosticism regarding metaphysics, Holden claims that Hume should be absolved of any inconsistency on the grounds that his ‘arguments concerning the intrinsic character of the first cause of all would apply equally to any other unknown being: they claim no particular insight into the distinctive intrinsic nature of the divinity’ (143). Whether this is an adequate reply lies beyond the scope of this short review. It seems fair, however, to note that the issue is not whether, in Holden's words, such an inference avoids ‘the overreaching speculation exemplified by core natural religion’ (143), but whether Hume's epistemology allows him to draw such a probabilistic inference regarding the divine nature. Can Hume consistently make any claims, however modest, concerning the intrinsic nature of a first cause?

Subsequent to his laying out what he understands to be Hume's arguments for strong moral atheism based on Hume's account of morality, Holden goes on to discuss Hume's treatment of the problem of evil, and his account of the implications of determinism for God's moral status. With regard to the problem of evil, he discusses Hume's treatment of three forms of the problem: the logical form, the inferential form, and the evidential form. Holden takes Hume to endorse the logical form of the problem, namely the claim ‘that the deity cannot be both infinitely powerful and good in any humanly recognizable sense’ (152). Holden does not feel that Hume is much interested in the logical form of the problem, however, since for Hume the hypothesis of an infinitely powerful being is a non-starter not worthy of serious philosophical consideration. Of more interest to Hume were attempts to argue from the nature of the world to the moral attributes of its original cause. Holden notes that Hume argues for the impossibility of such a project on two fronts. First, Hume maintains that, if asked to make a prediction, we would not have expected the deity to produce a world such as this, and second, that the universe is so unique we can draw no analogies based on our experience of other causes. (Holden does not raise the issue of whether Hume can consistently urge both these claims.) Holden also finds a version of the evidential form of the problem of evil in Hume, but argues that he does not seriously endorse it, but rather uses it to parody traditional core natural theology.

Hume's embrace of determinism, namely ‘his view that exceptionless natural laws govern all human thought and behavior, and that all events, including all human actions, fall out according to an unfolding deterministic causal system’ (181) suggests two arguments, each of which ‘purports to take us to a (different) negative conclusion concerning God's moral status’ (181). The first, which Holden views as ‘authentically Humean’, is that if God is the remote determining cause of all human actions, then none of those actions can violate its rights or prerogatives. Given that it determines all that we do, it cannot make moral claims on our behaviour and we cannot owe it moral duties. The second argument from determinism suggests that, since God determines all that happens, God is the author of sin. Holden does not see this argument as ‘authentically Humean’, but rather as a deliberately provocative ad hominem device used by Hume to point out internal tensions in Christian theology.

It is something of a puzzle why Holden includes these two chapters on Hume's treatment of the problem of evil and the theological implications of determinism. In both cases, Holden is at pains to point out that they provide no evidential support for strong moral atheism. Holden's case for reading Hume as a strong moral atheist rests, therefore, entirely on his claim that such a position ‘simply draws out the implications of … [Hume's] sentimentalism and moral psychology for the moral standing of any ultimate cause or organizing principle’ (214). He may well be right in this. It is scarcely surprising that if one articulates a totally naturalistic account of moral psychology that sees moral judgments as solely the expression of human emotion, as ‘simply a raw feeling with no intrinsic, or necessary intentional or object-directed character’ (54), one reaches the conclusion that God is beyond the pale of morality.

A bigger question is whether Hume was consistent in his views. Holden claims not only that Hume's moral psychology leads to strong moral atheism, but that Hume, although he never explicitly asserted strong moral atheism, was aware of this. Thus Holden attempts to explain away what he describes as Hume's ‘more tub-thumping statements of mitigated skepticism’ (214) which seem to tell against Holden's view that Hume consciously endorsed strong moral atheism. I am not as familiar with the corpus of Hume's works as Holden but, despite his careful handling of texts, I found myself at times feeling that he was trying for too much. It is one thing to suggest that Hume's moral psychology has implications for natural theology, quite another to insist that, despite never mentioning those implications, Hume was clearly aware of them, and that remarks which suggest otherwise are instances of tub-thumping or rhetorical grandstanding.

Whether or not one embraces the claim that Hume should be read as consciously advocating strong moral atheism, Holden provides a detailed argument that Hume's account of morality is inconsistent with any view of God as possessing moral attributes. In seeking to integrate Hume's theories concerning moral psychology with his well-known critiques of natural theology, this monograph provides an important exercise in reading Hume holistically. One need not think that Hume is invariably consistent in his views, or even to agree with the central tenets of Hume's philosophy, to see the value in this approach.