Linguistic hypercorrectness versus colloquial use in Latin at first century BC and afterwards

Abstract I discussed a passage in the Rhetorica ad C.Hernnium (82 BC) where two versions of the same talk are delivered ([1a] and [1b]), the former in a colloquial style, the second in a worse, almost “vulgar” Latin. The difference appears also in the two constructions (1a) coepit defricari (instead of *coeptus est defricari which is usually employed by Cicero, Caesar, Livy) and (1b) praesente multis (instead of *praesentibus multis). The former construction is correct but less used by classical authors; the second is not correct and appears as a kind of “vulgar” Latin. However, the Auctor ad Herennium knew very well the differences, because he uses coeptum est dici (4.30.41) and contra intercedentibis collegis (1.12.21). It seems therefore that the Auctor established a kind of barrier inside of which the language, albeit low, was correct, outside incorrect, and acknowledged that inside/outside was ruled by gradation and frequency of use and social condition.