Richard Swinburne’s new book “is concerned solely with the central core of theistic belief, that God exists, that there is a God. It is not concerned primarily with whether this belief is true or with whether \ye can know it to be true, but with the prior question of what it means and whether it is coherent.” (p. 1) The discussion begins with a general analysis (Chapters 4-6) of the way in which talk of God is to be understood, whether it is analogical, propositional and so forth. There then follows (Chapters 7-15) a consideration of what Swinburne takes to be involved in the claim that there is a God. Swinburne concludes that theism may be coherent, but that a certain understanding of divine attributes is necessary to sustain this conclusion. In particular, God’s eternity cannot be equated with timelessness, and both ‘omnipotent’ and ‘omniscient’ need to be understood in more narrow a fashion than is sometimes the case. Swinburne also maintains that the coherence of theism may only be properly affirmed where there is good, inductive evidence for the truth of theism. There is no direct proof of coherence. Consequently, “those theists who claim to believe that there is a God ‘by faith’, in a sense of the latter expression which entails that they do not have good inductive grounds for this belief, will, if I am right, have to face the consequence that they do not have good grounds for believing that the claim which they make is a coherent or logicallly possible one.” (p. 296)

So much for the overall argument of the book. What about the cogency of its details? It seems to me that Swinburne’s whole discussion is spoiled by a fundamental error of approach, a pervasive wrong move. On this I shall comment presently.