A question which arises from the presence in the MS of so many corrections concerns the degree to which the scribe's revision is 'roving'--that is, somewhat random, without concern for particular narrative sequences or themes. In an attempt to shed some light on the problem, I have produced this graph of the number of lines revised on each folio side of MS Ashmole 33. My judgement of what entails a revision is necessarily subjective, but my general policy has been to ignore scribal corrections of obvious currente calamo errors. Lines existing as palimpsests are counted, as are lines added in all margins.
The graph confirms Herrtage's observation that "the latter part of the poem... is much corrected" (p.xvi), and reinforces his suggestion that the tail-rhyme section was "possibly written off at once, and not copied from a previous draft" (p.xvi). While almost every folio side has received attention from the scribe (confirmation that his revision was to some extent roving), the extreme jaggedness of the graph shows that many sides received markedly more revision than those adjacent and were obviously the objects of special attention; only a more detailed study of the precise nature of revisions on each of these folios--something beyond the scope of the present investigation--will show to what degree those peaks represent additional roving or more specific revision.
Even so, larger scale 'groupings' of high (or no) revision do suggest the existence of a distinct revision policy. One large span of text in the couplet section (fols 33-41) shows virtually no revision at all and suggests the scribe's general satisfaction with the state of the translation there. Certain spans of narrative show revision well beyond the standard deviation above the average per folio side for each verse form (for the couplet section the standard deviation above average amounts to approximately six revisions per side; for the tail-rhyme section the figure is approximately twelve). These heavily-revised sections are fols 1-12 (the account of Oliver's duel with Ferumbras and his subsequent capture by Saracens); fols 27-33 (the capture of Floripas's bower); fols 49-51 (how Richard miraculously avoids the Saracens on his mission to France): and fols 61-73 (the final Saracen assaults on the bower and the final great battle). The average amount of revision in each of the four groups is generally greater than that of the preceding group, suggesting perhaps that, since each group involves key events in the story, the scribe was concerned with a suitable sense of progression toward a climax; certainly, his revision has in itself a kind of climactic movement.
To conclude, the revision appears to be both roving and specific; there is a certain amount of background roving revision interspersed with apparently more purposeful revision confined to discrete sections of the narrative.