This is the first paper I've seen come out of this year's IQSA conference, and it's pretty interesting. It relates to a key debate over whether the Qur'an was composed in an 'oral' context, which is a complicated issue.
http://www.academia.edu/9490706/Retelling_the_Tale_A_Computerised_Oral-Formulaic_Analysis_of_the_Qur_anThe scholar argues that his computer analysis shows extensive use of formulaic verses in the Qur'an, and that this use of formulaic verses is indicative of oral composition. I disagree generally with that conclusion, because it involves comparing a single *oral composition* against a single *written composition*. But the revisionist picture of the Qur'an's composition argues that it was written in stages, cobbled together, patched together by editors who added material, fixed transitions, and changed the rhymes. That patching together was surely done by the extensive use of stock phrases with slight variations. For example, Deroche identifies several 'short verses' that were added by scribes to fix the Qur'an's rhyme scheme, and only later were those new short verse divisions fixed to make it appear like a single longer verse.
So it is hard to see how the argument for orality here would be any different; the revisionist account itself presumes extensive use of formulaic phrases.
One point of crucial interest. You may recall my argument that the original ur-Qur'anic material likely resembled the short Surahs at the end of the Qur'an. In my view, similar archaic fragments were later welded into extensively reworked larger Surahs by the composer-scribes who created the long Medinan Surahs over the decades after Mo's death. The computer analysis supports that view (I think) because it shows that the 'later' Medinan surahs use FAR more formulaic elements than the 'earlier' Meccan surahs, precisely the opposite of what the traditional narrative about Qur'anic composition would suggest, but precisely what my view (that the Medinan surahs reflect far more intensive later scribal reworkings, fusings, and alterations, using stock phrases and patchwork) would expect. From the article:
"A third fascinating feature that emerges from the computer analysis is how those suras that have traditionally been labelled “Meccan” appear to be considerably less formulaic than those traditionally labelled “Medinan”. Overall, there is a significant difference in how those two groups of suras use formulaic language, as the following table illustrates:
“Meccan” 39.34% 23.92% 15.44% “Medinan” 55.01% 37.54% 25.50%
Space does not permit us here to explore what might account for these statistically significant differences: suffice to say that even if one were to dispute that the formulaic patterning the computer is uncovering is a sign of primary orality, it clearly is a feature whose use varies across the different periods of qur’anic composition."
Yes. One might well ask why the longer + later surahs are far more formulaic and repetitive than the early + shorter surahs!
Topic: Computer Analysis of Qur'anic Orality