As Professor Williams points out, there is no shortage of recent translations of the Confessions into English. Williams's initial motive for adding to these was to supply a translation that would be ‘sensitive to the philosophical content’ of the work (p. xvii). Thus he makes use of ‘more connective words (such as “for”) than is usual in contemporary English’ so that the ‘conceptual connections are made evident’, and uses the same English word to translate the same Latin word ‘in passages where that word, or the concept it represents, is being subject to analysis’ (p. xxx). But it is not only at this micro-level that Williams gives careful attention to the philosophical content. There are, in the introductory material, and in notes throughout, carefully thought-out discussions of the overall structure and purpose of the work as a whole, and of individual books and sections within it. However, this concern for a philosopher's translation, though it is never abandoned, came to be rivalled and surpassed by a concern for the way Scripture functions in Augustine's thought and in his text. In a finely argued section of the Introduction, Williams explains how he came to the decision not to mark Augustine's quotations from or allusions to scriptural passages within the text by quotation marks or italics. ‘Scriptural language is [Augustine's] language, the Psalmist's voice his voice… God is not merely the object but the source of praise: God gives the need for praise and gives also the will and the words by which that need is fulfilled’ (p. xxiif). In consequence, the only indications of Augustine's dependence on Scripture are to be found in the marginal references and, where Augustine's scriptural text differs from what the modern reader might be familiar with, in more discursive footnotes. Williams has a good ear for Scriptural echoes, and if he is right, as I think he is, to hear an echo of the fellow citizens of Ephesians 2:19 at IV.15.26 then there can be no doubt but that civibus meis must refer to fellowship in the city of God.

Another consequence of Williams's emphasis on intertextuality, and given the initial concern for Augustine's philosophical argument, an unexpected one, is that there are no references to philosophical sources or influences on his thought. The reason for this is that ‘On nearly every page of the Confessions, and many times on most pages, we find the words of Scripture: quoted exactly, slightly adapted, developed, extended, the object of comment, meditation, argument, and analysis. Nowhere, not once, do we find such a treatment of the works of the Platonists; and when Augustine does tell us what he learned from the Platonists, crucially important though it was, he tells us in the words of Scripture. So I have given credit where Augustine thinks credit is due, and kept silent where Augustine thinks it best to keep silent’ (p. xxxii). Although Williams's treatment of Augustine's use of Scripture is a clear advance on more traditional approaches, I think an occasional Platonic, or neo-Platonic, reference might have been useful to readers unfamiliar with the background, however hackneyed the reference might be; when Augustine speaks of having had his ‘back to the light and [his] face to the things on which it shone’ (IV.16.30), for example, or when he speaks of the ‘One’ and the ‘Many’, especially if these are to be given initial capitals (as they are at II.1.1, though not at XI.29.39).

The Introduction provides brief discussions of Augustine's life, thought and influence; the structure of the Confessions; intertextuality and Augustine's use of Scripture, and notes on the translation. These lively and engaging discussions leave one wishing that Williams had given himself a little more scope and his readers a little more food for thought. At the beginning of each of the thirteen books of the Confessions Williams supplies two brief introductory paragraphs. The first gives a general introduction to the material covered in the book; the second breaks this down into a dozen or so blocks of material.

The translation is careful, clear, vigorous and, generally speaking, accurate. But Williams is sometimes inclined to tinker with Augustine's use of metaphor and poetic imagery. His translation of Book III opens thus: ‘I came to Carthage, and all around me was the noisy bubbling and sizzling of disgraceful loves’. This translation certainly has an impact, but ‘noisy bubbling and sizzling’ is rather heavy freight for strepebat to bear on its own, and one rather misses the frying pan, the sartago, that was the seat of it all. Curiously enough, it does pop up, a few sections later, but as a cauldron (the word preferred for sartago by some earlier translators), where Williams has Augustine asking why it is that the stream of friendship runs ‘into a bubbling whirlpool of pitch, a monstrous cauldron of foul lusts, its course changed and its path diverted, twisted and cast down from heavenly serenity by its own impetus’ (III.2.3). But Augustine's own imagery here stays with that of the running stream; which runs not into a whirlpool but into a torrent (torrentem) of boiling pitch; not into a cauldron but into massive, heaving seas (aestus inmanes).

Williams tells us that he ‘wanted the language of the translation to be as musical as possible’ and that he had ‘taken great care over the rhythm and cadence of the language’ (p.xxx). He has certainly succeeded in this, but I could find no explanation of why in several places Augustine's prose has been set out in verse form (IX.10.25; X.27.38-28.39; 43.69-70; XII.10.10; 14.17: XIII.13.14).

This translation has been a painstaking labour of love. It will provide its readers with the opportunity for a fresh, thoughtful, and challenging engagement with Augustine's great work.