The three sets of texts that were explored all have a distinct status within the Afrikaans literary system. Niggie (2002) was awarded the prestigious Hertzog Prize for literature in 2004 and occupies a prominent position as a canonised text. Ingrid Winterbach, as the author, resides in the highly canonised strata of the Afrikaans system, rendering both her and the source text rich in symbolic capital. For Vaselinetjie (2004), Anoeschka Von Meek received the prestigious MER Prize for youth literature, the Jan Rabie/Rapport Prize for an innovative Afrikaans novel and the M-Net prize for Afrikaans texts in short format, all in 2005. The source text is a canonised youth ‘crossover’ novel prescribed at secondary school level in South Africa. A film version was released in 2017. Christiaan Bakkes is a seasoned traveller and game ranger in Africa, but not a canonised author nor a winner of any literary prizes. As deduced from the correspondence between the production agents, the target readers of his anthology of short stories In bushveld and desert are mainly tourists to Southern Africa.

In this case study we investigated the respective agents operative in the translation process. Among the agents, the common denominator was the translator, working with an array of other agents (including the source text authors) acting as revisers, editors and proofreaders under the guidance of three different project managers.2 We explore which agent shows the strongest ability to have their preferred translation selection as the final option to go into the printed target text, by analysing the way in which the discussions (or distinctive discursive lines) around these options are presented.

The agents operating in the translation projects studied here are mostly female. Only two male agents were operative, namely the author of In bushveld and desert and the editor for Niggie. The female agents are all between 45 and 65 years of age, have the same level of expertise, and are highly regarded by their various project managers. The translator has a Master’s degree in translation studies and the other agents, with the exception of the author of Vaselinetjie, all have postgraduate language schooling in literary studies and/or linguistics. Furthermore, being South African projects, all agents have a thorough command of both source text language (Afrikaans) and target text language (English). Therefore, the agents are all equal regarding their language abilities.

In our investigation of social power relations between the various agents, we focus entirely on personal power as revealed in the “interactional form of discourse control” (Van Dijk 1993:254). Our dataset’ consists of the agents’ e-mail correspondence (nearly 30,000 words in total). Our method resides within the domain of an archival methodology as introduced by Munday (2013), with the dataset as the archive on which we drew.

The framework of critical discourse analysis (CDA) provided the methodological tools for the analysis. In his work on CDA, Van Dijk (1993:283) points out the pivotal “role of discourse in the (reproduction ... of dominance”. This relationship can be traced back to relations between discourse and cognition, with discourse structures playing the mediating role. When viewed from this perspective, discourse structures can be seen as “the ‘symbolic’ reproduction of dominance, which plays out as social cognition taking on the form of “mind management” with consequences for understanding, attitudes and plans. For Van Dijk (1993:254) “the management of the mind of others in one’s own interests” is the modern form of power.

Practically, this means that macro-level notions such as dominance as a manifestation of power (with the implication of mind management) can be related to micro-level notions such as text (Van Dijk 1993). In a case study by Van Dijk (1993:270) on the parliamentary production of racism to “systematically examine the many textual and contextual properties of the exercise of dominance”, the focus therefore falls on those textual properties “that most clearly exhibit the discursive properties of the exercise of dominance”. The identification of symbolic reproductions of dominance therefore allows us to map power relations within the textual production process as macro-level notions.

For our study, we turned to microstructures or textual properties that reflect the exercise of power. Following Van Dijk (1993), who lists various properties (discourse phenomena) in this regard {inter alia specific lexical terms, pronoun use, speech acts and style), we designed our study to enable us to find specific discursive elements that could be interpreted as symbolic reproductions of dominance. Through the identification and further analysis of these reproductions, we could map the power formations in the three contexts.

Using a mixed-methods design, we sought to identify these discursive elements in the agents’ discourse, and set out three classes of such elements in the dataset to be “visually tracked and recorded” (Scocchera 2015:168) with the use of the analytic tool Atlas.ti.8. These classes are lexical chains, speech acts and modifying adverbs. We started from the assumption that a particular (macro-level) power stance would be marked by specific symbolic reproductions of dominance (linguistic choices) by the agent. Atlas, ti’s search function was applied to the dataset in order to identify and code the items mentioned above in order to generate a frequency count for each individual marker. After this stage of quantitatively extracting the most frequently used markers, we qualitatively investigated their individual occurrences in context, in order to be able to trace the construction of power relationships in the ongoing text production process.

First, we compiled a list of possible lexical chains, i.e. personal pronouns and verbs, based on a preliminary reading of the dataset. The chains we searched for were:

  • • “I think/don’t think”; “it seems” (indicating less assertiveness, and thus a position of lesser dominance or power);
  • • “You should” (indicating more assertiveness, and thus a position of greater dominance or power);
  • • “You must” (indicating even more assertiveness, and thus a position of even greater dominance or power).

Second, the qualitative search for speech acts meant that segments of text were coded as suggestions, requests, questions and imperatives. In this regard we followed Van Dijk (1993:250), who states that “we may assume that directive speech acts such as commands or orders may be used to enact power”. We therefore believe that an imperative is indicative of a position of greater power, while a suggestion implies a position of lesser power. We realised, however, that the coded segments of text would need to be subjected to a further contextual analysis, since, depending on its pragmatic environment, an utterance (for example, classified as a question) could be interpreted as a representation of either a more or a less assertive stance.

In the third place, adverbs functioning as sentence modifiers were coded after which a frequency count was generated. The adverbs identified in our archive were:

  • • moderators: bietjie (a little), mos (you know), darem (after all/some-how), Hewer (rather), eintlik (actually/rather);
  • • intensifiers: bepaald/beslis, defbiitief (definitely/surely), ongetwyfeld (without a doubt), absoluut (absolutely), regtig (really).

We assumed that moderators would be associated with lesser dominance, while intensifiers would in turn be associated with greater dominance.

< Prev   CONTENTS   Source   Next >