Inerrancy and belief structures

J.P. Moreland addresses inerrancy in his article The Rationality of Belief in Inerrancy in the Trinity Journal 7.1 (Spring 1986): 75-86. He has some interesting thoughts on why occasion difficulties with the text should not lead one to abandon inerrancy. Rather a less plausible inerrant reconstruction of a difficult text may be rationally preferred over a more plausible errant reconstruction if one otherwise has good reason to believe in inerrancy

An interpretation of a problem passage that harmonizes it with another text, and thus preserves inerrancy, may not be as rational as an interpretation that admits an error in the text, if one only considers this particular problem in isolation from other epistemically relevant considerations. But if one considers the depth of ingression of inerrancy as well, then the rationality of preserving belief in inerrancy in one’s noetic structure (as opposed to denying it and having to readjust a large part of one’s structure) can justify suspending judgment or believing a harmonization.

That a person does not change his entire belief structure every time he gains new information is not irrational

The defender of inerrancy is not being irrational in calling for suspension of judgment, etc. because of the depth of ingression of belief in inerrancy. At the very least, critics of the rationality of inerrancy should consider this before they criticize as irrational the various attempts to harmonize passages and the like.

The article as more useful for its arguments about the nature of belief as it is for its defence of inerrancy.

Such refusal to modify one's belief with minor unexplainable facts does not mean that one's belief can never change. We know that our knowledge is limited. Contrary evidence may appear contrary on the surface, yet we may lack even further information which in fact harmonises a rogue fact with our belief system. But enough contrary evidence can eventually modify our belief structure. This seems similar to Thomas Kuhn's argument about paradigm shifts in science. Moreland draws some parallels with scientific thinking.

So how much contrary evidence is required to overturn a belief structure. Well it depends on how central such a belief is, not so much in terms of conviction, but how many other beliefs are affected. Moreland addresses this earlier in the article, but also the problem that the criteria for rejection of held beliefs is not well defined by individuals.

It could be objected that nowhere have I stated any criteria for knowing when there would be enough problems with inerrancy to justify giving it up. Thus, I cannot rationally claim to know that inerrancy is true.

...As an example, consider a puzzle from the ancient Greeks, known as the sorites problem. Given a small heap of wheat, can I get a large heap by adding one grain? It seems not, for how could one go from a small to a large heap by merely adding one grain. But then it seems that one could add grains of wheat to a small heap and never reach a large heap.

Consider another puzzle. If one gradually changes the shade of a color from red to orange, can one tell when the color changes from red to orange? Probably not. But in the absence of such a criterion, how can I know when I see red or orange?

The problem with both puzzles is this: they assume that in the absence of clear criteria for borderline cases, one cannot have knowledge of clear cases. Without being able to judge when the heap becomes large, I can never know that it is large. Without being able to judge when the color changes to orange, I can never know that it is orange. But the fact is, I can know a large heap or an orange color even if I have no criteria.

I am not dismissing criteria altogether. Indeed, they are important in an overall theory of rationality. But I do not need criteria in all cases to know something.

I would add it depends on whether the new belief system which explains the rogue data also explains data in the old system. Which system if held to be true makes the most sense of all the data and has the least anomalies.