A thought

Just a thought, but I wonder if textual corruptions like those mentioned in the previous post, are known to be fewer in interrogative sentences or clauses? Given that texts were read aloud, the phonetic stress I presume were caused by a question could have perhaps aided copyists in their scribal accuracy. Why? Because of the psychological principle that things that are less usual, and, paradoxically, more detailed (here at a phonetic level) are more accurately dealt with by human memory. It would make an interesting study to compare the frequency of textual corruptions in relation to certain potential mnemonic features within a text.

Or have textual discrepancies already been analysed in such a fashion?