Deconstruction – the idea fails simply because of how it’s named. For something to be deconstructed it must first be constructed. That is to say, before something can be torn down and rebuilt, it must be proven to be something that was built up to begin with and not inherent to the system.
We are walking along in the woods and come across a hill and a building on that hill. If we want, we can tear down the building and rebuild it – but we still have to use the same materials. The hill, however, cannot be torn down (for the sake of the analogy, ignore the idea of blasting powder or earth moving machines – although the analogy still works because the land still exists though the hill does not, thus it is not properly deconstructed). We cannot deconstruct the hill because it wasn’t constructed to begin with – it is a natural portion of the world.
In epistemology, therefore, if it can be proven that there are certain beliefs that are a priori, or more importantly that are not constructed but basic beliefs, these beliefs are free from deconstruction. Furthermore, beliefs that rest directly upon the basic beliefs are likewise ‘safe.’ The further a belief is from its foundation, the more likely it is to be deconstructed.
It is akin to the ‘phone game,’ but modified. In the phone game one person whispers something into another person’s ear. That person, then, whispers what he thought he heard into someone else’s ear. Eventually, as things progress the original message is skewed. Proper basic beliefs, then, are similar. Though the basic belief is free from being deconstructed (since you can’t deconstruct what was never constructed), it’s predecessors do run the risk. For instance:
A (Basic belief) –> B –> C—> D
By the time we get to D there has been quite a bit of human interaction with the belief, thus one should look at the belief and – in a sense – deconstruct it. However, it is not proper deconstruction, because one would be evaluating D on its closeness to A rather than just blindly tearing it down. In other words, it is best to label it epistemic renovation than deconstruction. Assuming that God is a basic belief (this is an assumption, not something I will go into proving right now), all views of God – whether they be Christian, Mormon, Muslim, Hindu, etc – would be the “D’s” of the above example. Therefore, one must evaluate D and see if it properly compares with A (this is done through reasoning and, I believe, special revelation). One does not properly deconstruct D, but simply re-evaluates D to see if it follows from A.
Though there are plenty of defensive arguments against deconstructionism (i.e. “Deconstruction merely looks at man’s epistemic image and not the ontology of a belief”), this one is quite offensive. In order to validate deconstruction – that is, show how and why it is adequate and good in a certain situation – one first has to show that the knowledge being deconstructed was, in fact, constructed in the first place.