In this chapter, we analyse editorial intervention in modal usage in a corpus of written South African (SA) English texts across a range of registers. The corpus consists of unedited and edited versions of the same texts with metadata about the authors and editors, who are users of different subvarieties of SA English. We focus on two factors that play a role in editorial decision-making: the linguistic background of the editor in relation to that of the author; the role of codified norms, as reflected in grammars, usage guides and dictionaries. Results indicate that editors reduce the frequency of all modals across text types or author subvarieties, except for the modal may which is increased marginally, consistent with normative advice pertaining to its use for epistemic possibility meanings. Typical SA usage, such as the extended use of must, is not targeted for normative intervention. Editors tend to make fewer changes to texts written by Black SA authors, and intervene most with texts written by Afrikaans English bilingual authors. Native white SA editors make more frequent changes than Afrikaans English bilingual editors.

This chapter investigates the intersection between gender, writing and editing of published written texts and endonormativity in South African English. We focus on three sub-varieties within the South African context: two indigenous strands, Afrikaans English (AfrE) and Black South African English (BSAfE), and the settler strand, White South African English (WSAfE). We use multifactorial methods to analyse the effects of gender amongst a set of linguistic and extra-linguistic variables conditioning the genitive alternation across unedited and edited texts produced by AfrE, BSAfE and WSAfE authors. The results show that gender plays a minor role in conditioning the genitive alternation for both authors and editors, demonstrating that the genitive choices of men and women are conditioned in similar ways. As expected, linguistic factors play the greatest role in conditioning the genitive alternation. Our findings confirm recent investigations into cross-varietal and register differences and show that while the direction of the effect of linguistic factors is the same across sub-varieties and registers, the strength of these factors differs in certain sub-varieties and registers. Our findings also confirm recent findings regarding the genitive alternation in second-language varieties and suggest a possible substrate transfer effect, in especially BSAfE writing. Although editorial intervention introduces subtle shifts in preferences for the two constructions, this intervention mostly reinforces the choices of authors.