Can religious beliefs be called language-games? People who answer "no" worry that if religious beliefs are isolated language games, it becomes difficult to see why they should be cherished.
I. In his 'Lecture on Ethics,' Wittgenstein distinguishes between absolute (the good) and relative (a good chair) judgments of value. Belief in God has a point only if certain consequences follow, which seems to falsify its absolute character for some. But Phillips wants to show that there is a certain logic to religious belief: it is not the outcome of events which is to determine whether God is victorious, but faith in God which determines what is regarded as victory (125). Love of God thus determines what is important. People who deny these two points of view falsify the nature of religious belief.
II. Wittgenstein asked if the non-believer really contradicts the believer when he says that he does not believe what the believer believes. There must be a common problematic to assert contradiction. The reality of God thus cannot be assessed by resorting to some common measure. We should resist the craving for generality in analyzing religious belief. What is involved in belief is not "weighing evidence or reasoning to a conclusion" but the ways in which it influences and regulates a person's life. Wittgenstein shows how the matter does not come down to a common measure when he writes, "I think differently, in a different way. I say different things to myself. I have different pictures. It is this way: if someone said: 'Wittgeinstein, you don't take illness as punishment, so what do you believe?' - I'd say: 'I don't have any thoughts of punishment'" (129-30). Believing in something means putting one's trust in it, sacrificing for it, letting it regulate one's life. Belief in the Last Judgment is not a testable hypothesis, but nonetheless absolute in determining their thinking. We have to realize that there are situations in which there are blunders and those in which the word blunder doesn't really apply. For a blunder, belief in God would be too big to count.
III. So, to recap, religious beliefs have an absolute character but they are not testable and decidable by some common scientific measure. But this is not to say that religion belongs to the sphere of the "unsayable." Rush Rhees, in his essay on Wittgenstein's builders, says that at the same time as the builder's language is used on the job, it is also used at home and in other contexts. Similarly, the language of religious worship is not simply cut off from everything outside. If it were an isolated language game, it would be no more than an esoteric game. Religion would then simply be a neat set of rules, and it would be impossible to distinguish between genuine and sham worship. In short, religion must take the world seriously. If religious believers try to explain away suffering, one may accuse them of not taking it seriously. The meaning and force of religious beliefs depend in part on the relation of these beliefs with features of human existence other than religion (137).
A religious belief could be read as a testable hypothesis, in which case it could be proved a blunder. Or it could be read as having a different meaning. The mother placing a garland on the statue of the Virgin Mary may not think the garland's value is prudential but she may be venerating the birth of her child as God's gift, thanking him for it and contemplating the virtues of motherhood as found in the mother of Jesus (137-38). "The beliefs involved as not testable hypotheses, but ways of reacting to and meeting such situations" (138). From the cruder perspective, the protection of the child determines whether or not the act of bringing the child to the Virgin and the alleged holiness of the Virgin have been efficacious or not. From the proper perspective from which to understand religious belief, the holiness of the Virgin determines the nature of protection (139). In conclusion, it does not make sense to ask for a proof of the validity of religious beliefs (141).