When I was a Psychology student, I had a lecturer who told stories of his own early life as a young clinical psychologist. One story he told was of a psychotic patient who was under his care. This man was quite normal in other ways, but he believed that he (the patient) was dead. So one day my lecturer decided to try some cognitive therapy on him:
Lecturer: You think you're dead, yes? Well, do dead people bleed?
Patient: No, of course not. How could they?
Lecturer: (Sticking a pin in him) Well, how about that?
Patient: Good God! That's amazing! I was totally wrong!... Dead people do bleed!
Anyone who has tried to have an argument with a religious believer about the many contradictions and errors in religious texts will find themselves in sympathy with my lecturer. Each logical step forward seems to be met with another piece of twisted logic or wilful blindness. Typical examples can be found in many places on the web: for instance, at the Secular Web site or Ebon Musings where intelligent and dedicated people have given up hours of their time trying to pin down religious believers on the many contradictions in the Bible, only to be met by increasingly convoluted and unlikely twists of logic and inference. I don't want to disparage these people in any way, but I hope they realise that they will never convince their opponents, and the only ones to benefit from a debate of this kind are the onlookers. If there are any examples of a biblical literalist being convinced to recant by argument, I'm not aware of them.
Why is this? Why is it so hard to pin someone down to an agreed interpretation of what a particular phrase or sentence means? If it's any consolation to atheists, the problem is not just theirs: it besets the whole history of Western philosophy. Dozens of philosophers have stated what they considered were clear and definite answers to philosophical problems, only to have the whole thing misunderstood -- as they saw it -- in many different ways by their readers.
The fact is -- as Wittgenstein finally pointed out about fifty years ago -- language is not intended for debates. For pre-Darwinian philosophers like Descartes, who believed that language was designed and delivered to humans by a divine power, this must have been almost impossible to understand. Only when we recognise that language is a human invention, designed for a specific purpose, can we see that there are some things which language just can't do.
Here's a simple example: imagine there are two types of fruit tree that grow on Mars. They look completely identical, but one smells of oranges and the fruit is instantly fatal: the other smells of lemons and the fruit is sweet and healthy. A party of starving explorers gets in touch with you by radio and asks for your help in determining which fruits they can eat. The only problem is that they grew up on Venus and have never smelt either an orange or a lemon before.
How could you possibly help them? Other than saying 'It smells like...', is there any other way you could use language to convey the difference between orange-smell and lemon-smell? You can recognise it in an instant, but can you say it? Could you confidently give the explorers the instructions that would save their lives? In general, language is very poor at describing smells because that capacity has never been needed. Apart from a few specialised areas like wine tasting and perfumery, there is just no need for ordinary people to describe smells to each other, and so language has never developed that capacity.
The same is true of exegetical debates. Nothing in the evolutionary development of the human race has made it crucial to be able to distinguish infallible knowledge from strong belief, or sense-data from sensations, so debates about terms like these are bound to be loaded with misunderstandings and misinterpretations. Even the most careful 'natural language' philosopher or theologian will eventually stray into using terms which are manufactured for the purpose, and thereby spawn a different interpretation in the mind of every different person who hears or reads them.
But even 'ordinary language' can pose the same problems. "I have a tree in my garden" sounds like a simple statement, for instance, but someone who wanted to interpret it in such a way as to dispute its literal truth could find lots of ways to do so:

• It's not YOUR garden, it belongs jointly to you and your spouse.
• It's only three metres tall: it's not a tree but a large shrub.
• It's a bonsai tree, not a real tree.
• It's not IN the garden, but on the boundary with your neighbour.
• It's a tree, but a cut-down dead one from Christmas.
• It's only a metaphorical 'tree'.
• It's a real tree, but only a metaphorical 'garden'.

And of course I could have meant any of these things, and without corroborating evidence there is no way for you or anyone else to know for sure what I did mean. But all the corroborating evidence will also take the form of assertions, and any of these can be disputed in the same way, and so on and so forth...
Multiply this by the number of assertions in a book like the Bible, for instance, and it will become obvious that people will be able to reach agreement about what something actually says only when they are motivated to do so. Since philosophers make progress by disagreeing, and religious believers gain status by sticking to their beliefs no matter what, any project which tries to force people against their will to agree on what a written text says is inevitably doomed to failure. That is just not what language is for.
So by all means engage believers in arguments about the literal truth of their holy books, but don't expect to win, or even to have them concede that you have scored a single point. Make sure there are plenty of onlookers with open minds and you may achieve something, but be prepared for intense frustration as your opponent limbers up for some bizarre mental gymnastics. What is bleeding obvious to you may be a microscopic quibble to them, and vice versa.