The apparent failure of traditional and other individual arguments for theism has led some philosophers to construct cumulative cases. Such cases make use of the premises of various individual arguments, but they use them in combination. Whether a successful case of this sort can be constructed is a serious question with no obvious answer. It is certainly not ruled out by logic, since combining the premises of several relatively weak (or even worthless) arguments for some conclusion can in some cases yield a good argument for that conclusion. Nor is it ruled out by religion, since the success of such a case would eliminate neither the possibility of faith nor the possibility of nonphilosophical ways of knowing God. Notice also that the vast majority of well -supported theories in both science and everyday life are established by cumulative cases, not by a single argument or a single piece of evidence. Thus, if theism can (at least for the purposes of philosophical inquiry) be treated as a metaphysical theory, then the only way to determine whether a convincing cumulative case for it can be made is to actually construct and evaluate such cases.
How one might be inclined to structure a cumulative case for theism depends on what one believes are the weaknesses and strengths (if any) of non -cumulative cases for theism. If, on the one hand, one thinks that the problem with the best individual arguments for theism is that, while each establishes a part of the theistic hypothesis, each falls short of establishing the whole theistic hypothesis, then one might attempt a " distributive " cumulative case for theism. Such a case would assign to each of several individual arguments the task of establishing a different part of the theistic hypothesis. If every part of the theistic hypothesis is established by at least one argument, then such a case may succeed. If, on the other hand, one thinks that the problem with the individual arguments is that, while many of them provide some support for theism, none provides enough support to make theism probable, then one might attempt an " incremental " cumulative case for theism. Such a case would try to show that, when one adds together all of the support for theism provided by various individual arguments, the result is support for theism that is suffi ciently strong to make it probable.
In this chapter, I sketch and critically discuss one cumulative case for theism of each of these two types. Although I will refer to these two cases as " my " distributive case and " my " incremental case, this should not be taken to imply endorsement. The distributive case I sketch is similar in some respects but not in others to the distributive