Consider this Wittgensteinian line of thought about the doctrine of the resurrection of Jesus (similar things can be said about many other doctrines).[note 1] The way to understand what believers say when they confess "Jesus is risen" is to see what role that utterance plays in their lives, both there and then (say, in the liturgy) and more broadly in their lives. Further, the word "risen" in this sense has little currency outside religious contexts, so the notion of a "literal" meaning of the word outside of a religious context is problematic, and hence the role that the utterance plays in lives is determinative for meaning. It is difficult, therefore, to say what the disagreement between literalists and non-literalists about the resurrection is really about.
It seems to me that there is a fairly simple response to this line of thought. For the sake of argument, let's take on board the Wittgensteinian assumption that language gets its meaning from use, and hence to examine the meaning of an utterance one must see what role it plays in a life. But we must not forget that a significant part of the role that an utterance plays is found in its inferential connections with other utterances, both deductive and non-deductive.
For instance, the utterance "Jesus is risen" provides inferential support (one needs to judge on a case-by-case basis whether it's deductive or not) for such claims as "Jesus is not dead", "Jesus's skeleton isn't presently lying in the earth", "Jesus is alive", "People saw Jesus after his resurrection by means of photons reflected from his body", etc. This inferential support is at least as constitutive of the meaning of the utterance as is, say, the utterance's liturgical role. Moreover, the it is also a part of the role of the utterance that it is (non-deductively) supported by such claims as "Jesus's tomb is empty and his body is nowhere to be found", "Jesus was seen eating fish after he died", "Some of those who report Jesus as being risen knew what Jesus looked like", etc.
Perhaps unlike the confession "Jesus is risen", the claims I gave as inferentially connected with this confession make use of ordinary vocabulary and in ordinary ways. These claims are a part of our ordinary non-religious language games. We ask whether someone is alive or dead, whether someone's skeleton is buried in the earth, whether someone's tomb is empty, whether anyone saw someone eating fish, and so on. Now the Wittgensteinian anti-literalist will insist that the inferential connections between "Jesus is risen" and such ordinary claims do not do justice to the religious significance of "Jesus is risen", that for that significance one must pay attention to the liturgical and motivational role of the utterance in the lives of believers, and so on. But the literalist can take all that on board. There is more to the resurrection than is captured by the inferential connections with various ordinary non-religious claims. But there is no less: these inferential connections are an essential part of the constitution of the phrase's meaning, in a way that the non-literalist has a harder time accounting for. What connection, for the anti-literalist, is there between "Jesus is risen" and the fact that some of those who reported seeing Jesus knew what he looked like? But on a literalist reading, the inferential connection is clear.
This gives us both a way to characterize the disagreement—the literalist takes such inferential connections with ordinary claims to play a central part in the constitution of the meaning of "Jesus is risen" while the non-literalist does not—and a reason to favor the literalist reading, since the literalist is operating with a fuller collection of data.
At the same time, the literalist reading of "Jesus is risen" has significant flexibility, because the inferential connections are often probabilistic. The claim that x is risen makes probable that x was seen by means of photons reflected from x's body. But it does not entail it. It may be that x's body no longer reflects photons or that nobody has seen x. Thus, the literalist need not be committed to the claim that Jesus was seen by means of reflected photons, and might even deny it (there is a strong tradition that Jesus's body had some very special properties) but the literalist does need to see the confession "Jesus is risen" as of such a sort as to make the photonic claim a plausible inference. On an anti-literalist reading, however, the photonic claim does not seem to be a plausible inference. And that favors the literalist.
This gives us a literalism without a commitment to any such dubious thing as a "literal meaning".
But there is a lesson for the literalist, too. The literalist should not take the meaning of "Jesus is risen" to be exhausted by the inferential connections to ordinary claims. That, too, would be basing the meaning on merely partial data.