Leftow (1989). The doctrine of divine simplicity has been even more embattled, and
many (though not all) analytic theists would subscribe to the view that this doctrine has
not yet received a formulation that is sufficiently perspicuous to make it a serious
candidate for acceptance. Finally, there is the doctrine of necessary divine existence—
that God is a Necessary Being. Early on, the prevailing assumption was that Kant has
shown that existence is not a predicate and that the existence of no being can be logically
necessary. Most analytic theists, however, have come to reject this position, and to hold
that God's existence is indeed logically necessary, though this view is by no means
unanimous. Overall, considerable progress has been made in the philosophical accounting
for the divine attributes and the divine nature.
Even after the positivist embargo had been lifted, the theistic arguments suffered from
formidable difficulties, stemming historically from Hume and Kant. There was the ban on
necessary existence, which immediately invalidated the ontological argument and (if
Kant was to be believed) the cosmological argument as well. Arguably even more
important was the doctrine, common to Hume and Kant, that causation requires an
observable relation between phenomenal entities.
end p.428
This not only excludes the possibility of a causal argument for divine existence, but it
rules out the very possibility of a causal relationship between God and the world, such as
is implied in the doctrine of creation. Finally, there was the assumption that a successful
theistic argument (or “proof,” as they used to be called) must be one that is compelling
for any rational person who contemplates it—that it must proceed from premises known
(or readily knowable) to all, by means of inferences whose validity is evident to all.
Because arguments of this strength are seldom available for any philosophically
interesting conclusion, this assumption does a lot to make life easy for critics of the
theistic arguments.
All of these assumptions have been forcefully challenged in recent years. After providing,
early in his career, a refutation for the argument of Anselm's Proslogion 2, Alvin
Plantinga (1974) astonished himself by discovering an ontological argument, loosely
based on Proslogion 3, that is unquestionably valid. (Similar arguments were devised by
the Norman Malcolm [1960], a Wittgensteinian, and by the process philosopher Charles
Hartshorne [1962].) Plantinga himself admits, however, that his argument is “not a
successful piece of natural theology” (1974, 219), since its premise—that divine
necessary existence is logically possible—is itself in question and cannot be supported by
compelling arguments. Still, the ontological argument is back in play, which is a most
unexpected development in view of the situation just a few decades ago.
The Hume-Kant ban on causation by unobservables not only puts a crimp on theological
discourse; it also rules out the postulation of unobservable causes in science, and even (as
both Hume and Kant recognized) casts a shadow over realism concerning ordinary
physical objects. Once their arguments had been overturned (too long a story to be told
here), the way was open for a reexamination of the cosmological argument. An
impressive contribution along this line came from William Rowe's The Cosmological
Argument (1975, 1998), a study of Samuel Clarke's version of the argument based on the
principle of sufficient reason. Rowe is a nontheist and does not fully endorse the