The philosophical defense of theism has taken many different directions in recent years. The effort to produce strictly demonstrative theistic proofs has not been completely abandoned, but it has long since moved from centre stage. Some of the new modes of philosophical theism are quite ingenious, such as Alvin Plantinga’s effort to construe belief in God as a properly basic belief, i.e. one for which it makes no sense to demand evidence. However, despite its ingenuity, Plantinga’s effort has not been widely accepted. Other efforts, such as the Wittgensteinian and other forms of fideism, are less creditable and have been severely and, it seems, decisively, criticized.
For many theists, the most fruitful course appears to be to regard theism as something like a well-confirmed scientific hypothesis. This, for instance, is the course recommended by, John King-Farlow. Far from having been laid to rest by Hume; the “religious hypothesis” continues to receive attention from numerous and diverse thinkers. Some argue that the nature of the physical cosmos makes the hypothesis of God’s existence more likely than not. Others contend that theism provides the most reasonable account of the occurrence of certain phenomena.
In this thesis I examine and evaluate some of the more prominent efforts to establish theism as a well-confirmed explanatory hypothesis. My objective is to determine whether these efforts succeed or whether, on the other hand, construing theism as an hypothesis leads to its disconfirmation by rendering it vulnerable to the problem of evil.
The first chapter attempts to clarify the issue by employing a negative example, i.e. a paradigm case of how not to approach theism as a scientific hypothesis will be presented. The recent rise of “scientific” creationism provides an outstanding instance of how the attempt to confirm theism can descend into pseudoscience. The creationists attempt to propose two alternative “models”–evolution and divine creation–which they claim are exhaustive mutually exclusive accounts of the development of the physical cosmos and the origin of life. They then argue that evidence from geology, paleontology, physics, and other natural sciences tends to confirm the creationist “model” to a much higher degree than its alternative.
During the past four years, a number of noted scientists have written books detailing the fallacies, distortions, and inaccuracies found in creationist literature. This thesis makes an even stronger claim. I argue that, despite the tendency of many recent philosophers of science to think otherwise, a clear distinction can be drawn between science and pseudoscience. Further, I show how demarcation criteria can be developed that unambiguously vindicate evolutionary theory as scientific and relegate creationism to the pseudoscientific. In order to defend this view, a necessarily lengthy digression is made into certain issues in the philosophy of science. Chapter one then concludes that any attempt to follow the creationist’s lead by directly competing with a scientific theory will inevitably reduce theism to pseudoscience.
Much more sophisticated attempts to defend theism as a well-confirmed hypothesis are found in the writings of George Schlesinger and Richard Swinburne. Chapters two and three are devoted to Schlesinger and Swinburne respectively. Both are experts in confirmation theory and attempt to show how the canons and methods of that field can be used to reformulate and strengthen certain of the traditional theistic arguments.
Schlesinger begins by formulating and defending some basic principles of confirmation theory. He then applies those principles in an effort to show that certain features of the cosmos and human experience are better explained by theism than by naturalism. For instance, on the basis of naturalism intelligent life is no more likely to develop in the cosmos than not. Further, even if intelligent life does develop, naturalism gives no reason for thinking that such beings will be capable of moral goodness. Schlesinger argues that intelligent beings who are capable of moral goodness are much more likely to exist on the theistic hypothesis. Hence, he claims that the discovery in the actual world of intelligent moral agents provides evidence that confirms theism but fails to confirm naturalism.
After detailing these arguments and Schlesinger’s replies to anticipated criticisms, I offer an extended critique of those arguments and rebuttals.
Swinburne’s procedure is similar. He employs Bayes’ theorem to formulate an inductive version of the cosmological argument. Like Schlesinger, Swinburne presents theism and naturalism as rival hypotheses: Either the universe is the uItimate, uncaused existent (naturalism), or the universe was created by a transcendent being who is uncaused and who has existed eternally (theism). Given these alternative hypotheses, Swinburne argues that theism has a much higher a priori probability than naturalism. The reason is that theism is supposedly a much simpler hypothesis. That is, theism posits the existence of a single uncompounded being with very simple attributes. Swinburne claims that it is much more likely that such a simple being would exist uncaused than something as complex as the physical cosmos.
Swinburne’s cosmological argument has some appearance of plausibility, but I argue that it fails in a number of ways. Swinburne never succeeds in showing that the universe needs explaining or that there is any reason to regard theism more a priori likely than naturalism.
It has long been a popular argument of religious apologists that certain extraordinary events have occurred that can be explained only in terms of the miraculous acts of divine agents. The occurrence of such an event would therefore be strong evidence in favour of theism. Chapter four examines the epistemological issues involved in the effort to confirm the occurrence of miracles. One issue is whether an event can ever be known to be physically impossible, i.e. the sort of event that can never be explained in terms of a law of nature.
I claim that it is not clear that such an event could be identified since there seem to be no acceptable criteria for indisputable paradigms of such events. Further, even if such events are identifiable, there is still the Humean issue of whether there could ever be sufficient evidence for the occurrence of a miracle. I conclude that, while the occurrence of miracles cannot be disproven, neither can they be so well confirmed that a naturalist would have to admit their occurrence.
None of the arguments examined in the first four chapters succeeds in providing good evidence for theism as an hypothesis. The final chapter argues for the opposite thesis–that if theism is regarded as an explanatory hypothesis, it is disconfirmed by the existence of evils. An hypothesis-disconfirming version of the problem of evil is developed and possible theistic counters are considered.
One such counter, exemplified in the works of Plantinga and Nelson Pike, is to deny that any formal contradiction is generated by the problem of evil. I contend that the problem of evil is far too serious for such a minimal defense to suffice and that a plausible theodicy is called for.
Schlesinger and Swinburne attempt to provide such theodicies and the bulk of the last chapter is devoted to their arguments. I conclude that neither theodicy is successful and that the problem of evil disconfirms the theistic hypothesis.
My conclusion is therefore that it is fruitless to attempt the defense of theism with tools and methods of the sort employed in the corroboration of scientific hypotheses. I end with a few reflections on the nature of theism and naturalism. In particular, I am interested in why there is continued resistance to naturalism even after theism has been carefully and thoroughly criticized.
 See, for instance, J.L. Mackie, The Miracle of Theism (Oxford: Clarendon Press, 1982), chapt. 12.